Showing 1–50 of 302 results for author: Xie, P

FlashVLM: Text-Guided Visual Token Selection for Large Multimodal Models

Authors: Kaitong Cai, Jusheng Zhang, Jing Yang, Yijia Fan, Pengtao Xie, Jian Wang, Keze Wang

Abstract: Large vision-language models (VLMs) typically process hundreds or thousands of visual tokens per image or video frame, incurring quadratic attention cost and substantial redundancy. Existing token reduction methods often ignore the textual query or rely on deep attention maps, whose instability under aggressive pruning leads to degraded semantic alignment. We propose FlashVLM, a text guided visu… ▽ More Large vision-language models (VLMs) typically process hundreds or thousands of visual tokens per image or video frame, incurring quadratic attention cost and substantial redundancy. Existing token reduction methods often ignore the textual query or rely on deep attention maps, whose instability under aggressive pruning leads to degraded semantic alignment. We propose FlashVLM, a text guided visual token selection framework that dynamically adapts visual inputs to the query. Instead of relying on noisy attention weights, FlashVLM computes an explicit cross modal similarity between projected image tokens and normalized text embeddings in the language model space. This extrinsic relevance is fused with intrinsic visual saliency using log domain weighting and temperature controlled sharpening. In addition, a diversity preserving partition retains a minimal yet representative set of background tokens to maintain global context. Under identical token budgets and evaluation protocols, FlashVLM achieves beyond lossless compression, slightly surpassing the unpruned baseline while pruning up to 77.8 percent of visual tokens on LLaVA 1.5, and maintaining 92.8 percent accuracy even under 94.4 percent compression. Extensive experiments on 14 image and video benchmarks demonstrate that FlashVLM delivers state of the art efficiency performance trade offs while maintaining strong robustness and generalization across mainstream VLMs. △ Less

Submitted 23 December, 2025; originally announced December 2025.

Comments: Under submission

DreamPRM-Code: Function-as-Step Process Reward Model with Label Correction for LLM Coding

Authors: Ruiyi Zhang, Peijia Qin, Qi Cao, Pengtao Xie

Abstract: Process Reward Models (PRMs) have become essential for improving Large Language Models (LLMs) via test-time scaling, yet their effectiveness in coding remains limited due to the lack of meaningful step decompositions in code and the noise of Monte-Carlo-generated partial labels. We propose DreamPRM-Code, a coding-focused PRM that treats functions as reasoning steps using a Chain-of-Function prompt… ▽ More Process Reward Models (PRMs) have become essential for improving Large Language Models (LLMs) via test-time scaling, yet their effectiveness in coding remains limited due to the lack of meaningful step decompositions in code and the noise of Monte-Carlo-generated partial labels. We propose DreamPRM-Code, a coding-focused PRM that treats functions as reasoning steps using a Chain-of-Function prompting strategy to induce modular code generation, enabling PRM training and application analogous to mathematical reasoning tasks. To address label noise, DreamPRM-Code introduces a meta-learning-based correction mechanism that leverages clean final-solution unit-test labels and performs bi-level optimization to refine intermediate labels. Applying on test-time scaling, DreamPRM-Code achieved state-of-the-art performance on LiveCodeBench with 80.9 pass@1 rate, surpassing OpenAI o4-mini. △ Less

Submitted 16 December, 2025; originally announced December 2025.

Seedance 1.5 pro: A Native Audio-Visual Joint Generation Foundation Model

Authors: Team Seedance, Heyi Chen, Siyan Chen, Xin Chen, Yanfei Chen, Ying Chen, Zhuo Chen, Feng Cheng, Tianheng Cheng, Xinqi Cheng, Xuyan Chi, Jian Cong, Jing Cui, Qinpeng Cui, Qide Dong, Junliang Fan, Jing Fang, Zetao Fang, Chengjian Feng, Han Feng, Mingyuan Gao, Yu Gao, Dong Guo, Qiushan Guo, Boyang Hao , et al. (172 additional authors not shown)

Abstract: Recent strides in video generation have paved the way for unified audio-visual generation. In this work, we present Seedance 1.5 pro, a foundational model engineered specifically for native, joint audio-video generation. Leveraging a dual-branch Diffusion Transformer architecture, the model integrates a cross-modal joint module with a specialized multi-stage data pipeline, achieving exceptional au… ▽ More Recent strides in video generation have paved the way for unified audio-visual generation. In this work, we present Seedance 1.5 pro, a foundational model engineered specifically for native, joint audio-video generation. Leveraging a dual-branch Diffusion Transformer architecture, the model integrates a cross-modal joint module with a specialized multi-stage data pipeline, achieving exceptional audio-visual synchronization and superior generation quality. To ensure practical utility, we implement meticulous post-training optimizations, including Supervised Fine-Tuning (SFT) on high-quality datasets and Reinforcement Learning from Human Feedback (RLHF) with multi-dimensional reward models. Furthermore, we introduce an acceleration framework that boosts inference speed by over 10X. Seedance 1.5 pro distinguishes itself through precise multilingual and dialect lip-syncing, dynamic cinematic camera control, and enhanced narrative coherence, positioning it as a robust engine for professional-grade content creation. Seedance 1.5 pro is now accessible on Volcano Engine at https://console.volcengine.com/ark/region:ark+cn-beijing/experience/vision?type=GenVideo. △ Less

Submitted 23 December, 2025; v1 submitted 15 December, 2025; originally announced December 2025.

Comments: Seedance 1.5 pro Technical Report

EcomBench: Towards Holistic Evaluation of Foundation Agents in E-commerce

Authors: Rui Min, Zile Qiao, Ze Xu, Jiawen Zhai, Wenyu Gao, Xuanzhong Chen, Haozhen Sun, Zhen Zhang, Xinyu Wang, Hong Zhou, Wenbiao Yin, Bo Zhang, Xuan Zhou, Ming Yan, Yong Jiang, Haicheng Liu, Liang Ding, Ling Zou, Yi R. Fung, Yalong Li, Pengjun Xie

Abstract: Foundation agents have rapidly advanced in their ability to reason and interact with real environments, making the evaluation of their core capabilities increasingly important. While many benchmarks have been developed to assess agent performance, most concentrate on academic settings or artificially designed scenarios while overlooking the challenges that arise in real applications. To address th… ▽ More Foundation agents have rapidly advanced in their ability to reason and interact with real environments, making the evaluation of their core capabilities increasingly important. While many benchmarks have been developed to assess agent performance, most concentrate on academic settings or artificially designed scenarios while overlooking the challenges that arise in real applications. To address this issue, we focus on a highly practical real-world setting, the e-commerce domain, which involves a large volume of diverse user interactions, dynamic market conditions, and tasks directly tied to real decision-making processes. To this end, we introduce EcomBench, a holistic E-commerce Benchmark designed to evaluate agent performance in realistic e-commerce environments. EcomBench is built from genuine user demands embedded in leading global e-commerce ecosystems and is carefully curated and annotated through human experts to ensure clarity, accuracy, and domain relevance. It covers multiple task categories within e-commerce scenarios and defines three difficulty levels that evaluate agents on key capabilities such as deep information retrieval, multi-step reasoning, and cross-source knowledge integration. By grounding evaluation in real e-commerce contexts, EcomBench provides a rigorous and dynamic testbed for measuring the practical capabilities of agents in modern e-commerce. △ Less

Submitted 11 December, 2025; v1 submitted 9 December, 2025; originally announced December 2025.

Cryptanalysis of Gleeok-128

Authors: Siwei Chen, Peipei Xie, Shengyuan Xu, Xiutao Feng, Zejun Xiang, Xiangyong Zeng

Abstract: Gleeok is a family of low latency keyed pseudorandom functions (PRFs) consisting of three parallel SPN based permutations whose outputs are XORed to form the final value. Both Gleeok-128 and Gleeok-256 use a 256 bit key, with block sizes of 128 and 256 bits, respectively. Owing to its multi branch structure, evaluating security margins and mounting effective key recovery attacks present nontrivial… ▽ More Gleeok is a family of low latency keyed pseudorandom functions (PRFs) consisting of three parallel SPN based permutations whose outputs are XORed to form the final value. Both Gleeok-128 and Gleeok-256 use a 256 bit key, with block sizes of 128 and 256 bits, respectively. Owing to its multi branch structure, evaluating security margins and mounting effective key recovery attacks present nontrivial challenges. This paper provides the first comprehensive third party cryptanalysis of Gleeok-128. We introduce a two stage MILP based framework for constructing branch wise and full cipher differential linear (DL) distinguishers, together with an integral based key recovery framework tailored to multi branch designs. Our DL analysis yields 7, 7, 8, and 4 round distinguishers for Branch 1, Branch 2, Branch 3, and Gleeok-128, respectively, with squared correlations approximately 2 to the power minus 88.12, 2 to the power minus 88.12, 2 to the power minus 38.73, and 2 to the power minus 49.04, outperforming those in the design document except for the full PRF case. By tightening algebraic degree bounds, we further derive 9, 9, and 7 round integral distinguishers for the three branches and a 7 round distinguisher for the full PRF, extending the designers results by 3, 3, and 2 rounds and by 2 rounds, respectively. These integral properties enable 7 round and 8 round key recovery attacks in the non full codebook and full codebook settings. In addition, we identify a flaw in the original linear security evaluation of Branch 3, showing that it can be distinguished over all 12 rounds with data complexity about 2 to the power 48. We also propose optimized linear layer parameters that significantly improve linear resistance without sacrificing diffusion. Our results advance the understanding of Gleeok-128 and provide general methods for analyzing multi branch symmetric designs. △ Less

Submitted 4 December, 2025; originally announced December 2025.

Comments: 44 pages, 5 figures

MSC Class: 94A60

Pluggable Pruning with Contiguous Layer Distillation for Diffusion Transformers

Authors: Jian Ma, Qirong Peng, Xujie Zhu, Peixing Xie, Chen Chen, Haonan Lu

Abstract: Diffusion Transformers (DiTs) have shown exceptional performance in image generation, yet their large parameter counts incur high computational costs, impeding deployment in resource-constrained settings. To address this, we propose Pluggable Pruning with Contiguous Layer Distillation (PPCL), a flexible structured pruning framework specifically designed for DiT architectures. First, we identify re… ▽ More Diffusion Transformers (DiTs) have shown exceptional performance in image generation, yet their large parameter counts incur high computational costs, impeding deployment in resource-constrained settings. To address this, we propose Pluggable Pruning with Contiguous Layer Distillation (PPCL), a flexible structured pruning framework specifically designed for DiT architectures. First, we identify redundant layer intervals through a linear probing mechanism combined with the first-order differential trend analysis of similarity metrics. Subsequently, we propose a plug-and-play teacher-student alternating distillation scheme tailored to integrate depth-wise and width-wise pruning within a single training phase. This distillation framework enables flexible knowledge transfer across diverse pruning ratios, eliminating the need for per-configuration retraining. Extensive experiments on multiple Multi-Modal Diffusion Transformer architecture models demonstrate that PPCL achieves a 50\% reduction in parameter count compared to the full model, with less than 3\% degradation in key objective metrics. Notably, our method maintains high-quality image generation capabilities while achieving higher compression ratios, rendering it well-suited for resource-constrained environments. The open-source code, checkpoints for PPCL can be found at the following link: https://github.com/OPPO-Mente-Lab/Qwen-Image-Pruning. △ Less

Submitted 20 November, 2025; originally announced November 2025.

Comments: https://github.com/OPPO-Mente-Lab/Qwen-Image-Pruning

Steganographic Backdoor Attacks in NLP: Ultra-Low Poisoning and Defense Evasion

Authors: Eric Xue, Ruiyi Zhang, Zijun Zhang, Pengtao Xie

Abstract: Transformer models are foundational to natural language processing (NLP) applications, yet remain vulnerable to backdoor attacks introduced through poisoned data, which implant hidden behaviors during training. To strengthen the ability to prevent such compromises, recent research has focused on designing increasingly stealthy attacks to stress-test existing defenses, pairing backdoor behaviors wi… ▽ More Transformer models are foundational to natural language processing (NLP) applications, yet remain vulnerable to backdoor attacks introduced through poisoned data, which implant hidden behaviors during training. To strengthen the ability to prevent such compromises, recent research has focused on designing increasingly stealthy attacks to stress-test existing defenses, pairing backdoor behaviors with stylized artifact or token-level perturbation triggers. However, this trend diverts attention from the harder and more realistic case: making the model respond to semantic triggers such as specific names or entities, where a successful backdoor could manipulate outputs tied to real people or events in deployed systems. Motivated by this growing disconnect, we introduce SteganoBackdoor, bringing stealth techniques back into line with practical threat models. Leveraging innocuous properties from natural-language steganography, SteganoBackdoor applies a gradient-guided data optimization process to transform semantic trigger seeds into steganographic carriers that embed a high backdoor payload, remain fluent, and exhibit no representational resemblance to the trigger. Across diverse experimental settings, SteganoBackdoor achieves over 99% attack success at an order-of-magnitude lower data-poisoning rate than prior approaches while maintaining unparalleled evasion against a comprehensive suite of data-level defenses. By revealing this practical and covert attack, SteganoBackdoor highlights an urgent blind spot in current defenses and demands immediate attention to adversarial data defenses and real-world threat modeling. △ Less

Submitted 25 November, 2025; v1 submitted 18 November, 2025; originally announced November 2025.

Reducing Hallucinations in LLM-Generated Code via Semantic Triangulation

Authors: Yihan Dai, Sijie Liang, Haotian Xu, Peichu Xie, Sergey Mechtaev

Abstract: When generating code from natural language prompts, an LLM samples programs from a probability distribution, many of which might be incorrect. Sample consensus techniques - such as majority voting or validation against generated tests or specifications - aim to identify a correct program in the sample or abstain if none is valid. However, existing methods often fail to select a correct solution wh… ▽ More When generating code from natural language prompts, an LLM samples programs from a probability distribution, many of which might be incorrect. Sample consensus techniques - such as majority voting or validation against generated tests or specifications - aim to identify a correct program in the sample or abstain if none is valid. However, existing methods often fail to select a correct solution when its sampling probability is low, or when the problem permits multiple valid but non-equivalent solutions. Additionally, they often fail to abstain when no correct solution is present in the sample. To overcome these limitations, we introduce semantic triangulation, which transforms a programming problem in a way that non-trivially alters its semantics while preserving an exact, verifiable mapping between solutions before and after transformation. We theoretically establish that verifying consistency across such problem transformations increases confidence that generated programs reflect accurate generalization rather than spurious statistical correlations, enabling more reliable sample consensus and abstention. On the LiveCodeBench and CodeElo benchmarks, using GPT-4o and DeepSeek-V3 models, semantic triangulation increases reliability of generated code by 21% compared to the method that selects only high-confidence solutions with the probability threshold 0.5, while being able to pinpoint correct solutions at sampling probabilities as low as 0.14. Apart from that, it is also the only approach to consistently form true consensus on tasks with multiple valid but non-equivalent solutions. △ Less

Submitted 21 November, 2025; v1 submitted 15 November, 2025; originally announced November 2025.

MMA-Sim: Bit-Accurate Reference Model of Tensor Cores and Matrix Cores

Authors: Peichen Xie, Yang Wang, Fan Yang, Mao Yang

Abstract: The rapidly growing computation demands of deep neural networks (DNNs) have driven hardware vendors to integrate matrix multiplication accelerators (MMAs), such as NVIDIA Tensor Cores and AMD Matrix Cores, into modern GPUs. However, due to distinct and undocumented arithmetic specifications for floating-point matrix multiplication, some MMAs can lead to numerical imprecision and inconsistency that… ▽ More The rapidly growing computation demands of deep neural networks (DNNs) have driven hardware vendors to integrate matrix multiplication accelerators (MMAs), such as NVIDIA Tensor Cores and AMD Matrix Cores, into modern GPUs. However, due to distinct and undocumented arithmetic specifications for floating-point matrix multiplication, some MMAs can lead to numerical imprecision and inconsistency that can compromise the stability and reproducibility of DNN training and inference. This paper presents MMA-Sim, the first bit-accurate reference model that reveals the detailed arithmetic behaviors of the MMAs from ten GPU architectures (eight from NVIDIA and two from AMD). By dissecting the MMAs using a combination of targeted and randomized tests, our methodology derives nine arithmetic algorithms to simulate the floating-point matrix multiplication of the MMAs. Large-scale validation confirms bitwise equivalence between MMA-Sim and the real hardware. Using MMA-Sim, we investigate arithmetic behaviors that affect DNN training stability, and identify undocumented behaviors that could lead to significant errors. △ Less

Submitted 13 November, 2025; originally announced November 2025.

IterResearch: Rethinking Long-Horizon Agents via Markovian State Reconstruction

Authors: Guoxin Chen, Zile Qiao, Xuanzhong Chen, Donglei Yu, Haotian Xu, Wayne Xin Zhao, Ruihua Song, Wenbiao Yin, Huifeng Yin, Liwen Zhang, Kuan Li, Minpeng Liao, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Recent advances in deep-research agents have shown promise for autonomous knowledge construction through dynamic reasoning over external sources. However, existing approaches rely on a mono-contextual paradigm that accumulates all information in a single, expanding context window, leading to context suffocation and noise contamination that limit their effectiveness on long-horizon tasks. We introd… ▽ More Recent advances in deep-research agents have shown promise for autonomous knowledge construction through dynamic reasoning over external sources. However, existing approaches rely on a mono-contextual paradigm that accumulates all information in a single, expanding context window, leading to context suffocation and noise contamination that limit their effectiveness on long-horizon tasks. We introduce IterResearch, a novel iterative deep-research paradigm that reformulates long-horizon research as a Markov Decision Process with strategic workspace reconstruction. By maintaining an evolving report as memory and periodically synthesizing insights, our approach preserves consistent reasoning capacity across arbitrary exploration depths. We further develop Efficiency-Aware Policy Optimization (EAPO), a reinforcement learning framework that incentivizes efficient exploration through geometric reward discounting and enables stable distributed training via adaptive downsampling. Extensive experiments demonstrate that IterResearch achieves substantial improvements over existing open-source agents with average +14.5pp across six benchmarks and narrows the gap with frontier proprietary systems. Remarkably, our paradigm exhibits unprecedented interaction scaling, extending to 2048 interactions with dramatic performance gains (from 3.5\% to 42.5\%), and serves as an effective prompting strategy, improving frontier models by up to 19.2pp over ReAct on long-horizon tasks. These findings position IterResearch as a versatile solution for long-horizon reasoning, effective both as a trained agent and as a prompting paradigm for frontier models. △ Less

Submitted 10 November, 2025; originally announced November 2025.

Comments: https://github.com/Alibaba-NLP/DeepResearch

Towards Universal Video Retrieval: Generalizing Video Embedding via Synthesized Multimodal Pyramid Curriculum

Authors: Zhuoning Guo, Mingxin Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Xiaowen Chu

Abstract: The prevailing video retrieval paradigm is structurally misaligned, as narrow benchmarks incentivize correspondingly limited data and single-task training. Therefore, universal capability is suppressed due to the absence of a diagnostic evaluation that defines and demands multi-dimensional generalization. To break this cycle, we introduce a framework built on the co-design of evaluation, data, and… ▽ More The prevailing video retrieval paradigm is structurally misaligned, as narrow benchmarks incentivize correspondingly limited data and single-task training. Therefore, universal capability is suppressed due to the absence of a diagnostic evaluation that defines and demands multi-dimensional generalization. To break this cycle, we introduce a framework built on the co-design of evaluation, data, and modeling. First, we establish the Universal Video Retrieval Benchmark (UVRB), a suite of 16 datasets designed not only to measure performance but also to diagnose critical capability gaps across tasks and domains. Second, guided by UVRB's diagnostics, we introduce a scalable synthesis workflow that generates 1.55 million high-quality pairs to populate the semantic space required for universality. Finally, we devise the Modality Pyramid, a curriculum that trains our General Video Embedder (GVE) by explicitly leveraging the latent interconnections within our diverse data. Extensive experiments show GVE achieves state-of-the-art zero-shot generalization on UVRB. In particular, our analysis reveals that popular benchmarks are poor predictors of general ability and that partially relevant retrieval is a dominant but overlooked scenario. Overall, our co-designed framework provides a practical path to escape the limited scope and advance toward truly universal video retrieval. △ Less

Submitted 31 October, 2025; originally announced October 2025.

Tongyi DeepResearch Technical Report

Authors: Tongyi DeepResearch Team, Baixuan Li, Bo Zhang, Dingchu Zhang, Fei Huang, Guangyu Li, Guoxin Chen, Huifeng Yin, Jialong Wu, Jingren Zhou, Kuan Li, Liangcai Su, Litu Ou, Liwen Zhang, Pengjun Xie, Rui Ye, Wenbiao Yin, Xinmiao Yu, Xinyu Wang, Xixi Wu, Xuanzhong Chen, Yida Zhao, Zhen Zhang, Zhengwei Tao, Zhongwang Zhang , et al. (32 additional authors not shown)

Abstract: We present Tongyi DeepResearch, an agentic large language model, which is specifically designed for long-horizon, deep information-seeking research tasks. To incentivize autonomous deep research agency, Tongyi DeepResearch is developed through an end-to-end training framework that combines agentic mid-training and agentic post-training, enabling scalable reasoning and information seeking across co… ▽ More We present Tongyi DeepResearch, an agentic large language model, which is specifically designed for long-horizon, deep information-seeking research tasks. To incentivize autonomous deep research agency, Tongyi DeepResearch is developed through an end-to-end training framework that combines agentic mid-training and agentic post-training, enabling scalable reasoning and information seeking across complex tasks. We design a highly scalable data synthesis pipeline that is fully automatic, without relying on costly human annotation, and empowers all training stages. By constructing customized environments for each stage, our system enables stable and consistent interactions throughout. Tongyi DeepResearch, featuring 30.5 billion total parameters, with only 3.3 billion activated per token, achieves state-of-the-art performance across a range of agentic deep research benchmarks, including Humanity's Last Exam, BrowseComp, BrowseComp-ZH, WebWalkerQA, xbench-DeepSearch, FRAMES and xbench-DeepSearch-2510. We open-source the model, framework, and complete solutions to empower the community. △ Less

Submitted 4 November, 2025; v1 submitted 28 October, 2025; originally announced October 2025.

Comments: https://tongyi-agent.github.io/blog

AgentFold: Long-Horizon Web Agents with Proactive Context Management

Authors: Rui Ye, Zhongwang Zhang, Kuan Li, Huifeng Yin, Zhengwei Tao, Yida Zhao, Liangcai Su, Liwen Zhang, Zile Qiao, Xinyu Wang, Pengjun Xie, Fei Huang, Siheng Chen, Jingren Zhou, Yong Jiang

Abstract: LLM-based web agents show immense promise for information seeking, yet their effectiveness on long-horizon tasks is hindered by a fundamental trade-off in context management. Prevailing ReAct-based agents suffer from context saturation as they accumulate noisy, raw histories, while methods that fixedly summarize the full history at each step risk the irreversible loss of critical details. Addressi… ▽ More LLM-based web agents show immense promise for information seeking, yet their effectiveness on long-horizon tasks is hindered by a fundamental trade-off in context management. Prevailing ReAct-based agents suffer from context saturation as they accumulate noisy, raw histories, while methods that fixedly summarize the full history at each step risk the irreversible loss of critical details. Addressing these, we introduce AgentFold, a novel agent paradigm centered on proactive context management, inspired by the human cognitive process of retrospective consolidation. AgentFold treats its context as a dynamic cognitive workspace to be actively sculpted, rather than a passive log to be filled. At each step, it learns to execute a `folding' operation, which manages its historical trajectory at multiple scales: it can perform granular condensations to preserve vital, fine-grained details, or deep consolidations to abstract away entire multi-step sub-tasks. The results on prominent benchmarks are striking: with simple supervised fine-tuning (without continual pre-training or RL), our AgentFold-30B-A3B agent achieves 36.2% on BrowseComp and 47.3% on BrowseComp-ZH. Notably, this performance not only surpasses or matches open-source models of a dramatically larger scale, such as the DeepSeek-V3.1-671B-A37B, but also surpasses leading proprietary agents like OpenAI's o4-mini. △ Less

Submitted 28 October, 2025; originally announced October 2025.

Comments: 26 pages, 9 figures

ParallelMuse: Agentic Parallel Thinking for Deep Information Seeking

Authors: Baixuan Li, Dingchu Zhang, Jialong Wu, Wenbiao Yin, Zhengwei Tao, Yida Zhao, Liwen Zhang, Haiyang Shen, Runnan Fang, Pengjun Xie, Jingren Zhou, Yong Jiang

Abstract: Parallel thinking expands exploration breadth, complementing the deep exploration of information-seeking (IS) agents to further enhance problem-solving capability. However, conventional parallel thinking faces two key challenges in this setting: inefficiency from repeatedly rolling out from scratch, and difficulty in integrating long-horizon reasoning trajectories during answer generation, as limi… ▽ More Parallel thinking expands exploration breadth, complementing the deep exploration of information-seeking (IS) agents to further enhance problem-solving capability. However, conventional parallel thinking faces two key challenges in this setting: inefficiency from repeatedly rolling out from scratch, and difficulty in integrating long-horizon reasoning trajectories during answer generation, as limited context capacity prevents full consideration of the reasoning process. To address these issues, we propose ParallelMuse, a two-stage paradigm designed for deep IS agents. The first stage, Functionality-Specified Partial Rollout, partitions generated sequences into functional regions and performs uncertainty-guided path reuse and branching to enhance exploration efficiency. The second stage, Compressed Reasoning Aggregation, exploits reasoning redundancy to losslessly compress information relevant to answer derivation and synthesize a coherent final answer. Experiments across multiple open-source agents and benchmarks demonstrate up to 62% performance improvement with a 10--30% reduction in exploratory token consumption. △ Less

Submitted 28 October, 2025; originally announced October 2025.

WebLeaper: Empowering Efficiency and Efficacy in WebAgent via Enabling Info-Rich Seeking

Authors: Zhengwei Tao, Haiyang Shen, Baixuan Li, Wenbiao Yin, Jialong Wu, Kuan Li, Zhongwang Zhang, Huifeng Yin, Rui Ye, Liwen Zhang, Xinyu Wang, Pengjun Xie, Jingren Zhou, Yong Jiang

Abstract: Large Language Model (LLM)-based agents have emerged as a transformative approach for open-ended problem solving, with information seeking (IS) being a core capability that enables autonomous reasoning and decision-making. While prior research has largely focused on improving retrieval depth, we observe that current IS agents often suffer from low search efficiency, which in turn constrains overal… ▽ More Large Language Model (LLM)-based agents have emerged as a transformative approach for open-ended problem solving, with information seeking (IS) being a core capability that enables autonomous reasoning and decision-making. While prior research has largely focused on improving retrieval depth, we observe that current IS agents often suffer from low search efficiency, which in turn constrains overall performance. A key factor underlying this inefficiency is the sparsity of target entities in training tasks, which limits opportunities for agents to learn and generalize efficient search behaviors. To address these challenges, we propose WebLeaper, a framework for constructing high-coverage IS tasks and generating efficient solution trajectories. We formulate IS as a tree-structured reasoning problem, enabling a substantially larger set of target entities to be embedded within a constrained context. Leveraging curated Wikipedia tables, we propose three variants for synthesizing IS tasks, Basic, Union, and Reverse-Union, to systematically increase both IS efficiency and efficacy. Finally, we curate training trajectories by retaining only those that are simultaneously accurate and efficient, ensuring that the model is optimized for both correctness and search performance. Extensive experiments on both basic and comprehensive settings, conducted on five IS benchmarks, BrowserComp, GAIA, xbench-DeepSearch, WideSearch, and Seal-0, demonstrate that our method consistently achieves improvements in both effectiveness and efficiency over strong baselines. △ Less

Submitted 28 October, 2025; originally announced October 2025.

AgentFrontier: Expanding the Capability Frontier of LLM Agents with ZPD-Guided Data Synthesis

Authors: Xuanzhong Chen, Zile Qiao, Guoxin Chen, Liangcai Su, Zhen Zhang, Xinyu Wang, Pengjun Xie, Fei Huang, Jingren Zhou, Yong Jiang

Abstract: Training large language model agents on tasks at the frontier of their capabilities is key to unlocking advanced reasoning. We introduce a data synthesis approach inspired by the educational theory of the Zone of Proximal Development (ZPD), which defines this frontier as tasks an LLM cannot solve alone but can master with guidance. To operationalize this, we present the AgentFrontier Engine, an au… ▽ More Training large language model agents on tasks at the frontier of their capabilities is key to unlocking advanced reasoning. We introduce a data synthesis approach inspired by the educational theory of the Zone of Proximal Development (ZPD), which defines this frontier as tasks an LLM cannot solve alone but can master with guidance. To operationalize this, we present the AgentFrontier Engine, an automated pipeline that synthesizes high-quality, multidisciplinary data situated precisely within the LLM's ZPD. This engine supports both continued pre-training with knowledge-intensive data and targeted post-training on complex reasoning tasks. From the same framework, we derive the ZPD Exam, a dynamic and automated benchmark designed to evaluate agent capabilities on these frontier tasks. We train AgentFrontier-30B-A3B model on our synthesized data, which achieves state-of-the-art results on demanding benchmarks like Humanity's Last Exam, even surpassing some leading proprietary agents. Our work demonstrates that a ZPD-guided approach to data synthesis offers a scalable and effective path toward building more capable LLM agents. △ Less

Submitted 28 October, 2025; originally announced October 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

Repurposing Synthetic Data for Fine-grained Search Agent Supervision

Authors: Yida Zhao, Kuan Li, Xixi Wu, Liwen Zhang, Dingchu Zhang, Baixuan Li, Maojia Song, Zhuo Chen, Chenxi Wang, Xinyu Wang, Kewei Tu, Pengjun Xie, Jingren Zhou, Yong Jiang

Abstract: LLM-based search agents are increasingly trained on entity-centric synthetic data to solve complex, knowledge-intensive tasks. However, prevailing training methods like Group Relative Policy Optimization (GRPO) discard this rich entity information, relying instead on sparse, outcome-based rewards. This critical limitation renders them unable to distinguish informative "near-miss" samples-those wit… ▽ More LLM-based search agents are increasingly trained on entity-centric synthetic data to solve complex, knowledge-intensive tasks. However, prevailing training methods like Group Relative Policy Optimization (GRPO) discard this rich entity information, relying instead on sparse, outcome-based rewards. This critical limitation renders them unable to distinguish informative "near-miss" samples-those with substantially correct reasoning but a flawed final answer-from complete failures, thus discarding valuable learning signals. We address this by leveraging the very entities discarded during training. Our empirical analysis reveals a strong positive correlation between the number of ground-truth entities identified during an agent's reasoning process and final answer accuracy. Building on this insight, we introduce Entity-aware Group Relative Policy Optimization (E-GRPO), a novel framework that formulates a dense entity-aware reward function. E-GRPO assigns partial rewards to incorrect samples proportional to their entity match rate, enabling the model to effectively learn from these "near-misses". Experiments on diverse question-answering (QA) and deep research benchmarks show that E-GRPO consistently and significantly outperforms the GRPO baseline. Furthermore, our analysis reveals that E-GRPO not only achieves superior accuracy but also induces more efficient reasoning policies that require fewer tool calls, demonstrating a more effective and sample-efficient approach to aligning search agents. △ Less

Submitted 28 October, 2025; originally announced October 2025.

BrowseConf: Confidence-Guided Test-Time Scaling for Web Agents

Authors: Litu Ou, Kuan Li, Huifeng Yin, Liwen Zhang, Zhongwang Zhang, Xixi Wu, Rui Ye, Zile Qiao, Pengjun Xie, Jingren Zhou, Yong Jiang

Abstract: Confidence in LLMs is a useful indicator of model uncertainty and answer reliability. Existing work mainly focused on single-turn scenarios, while research on confidence in complex multi-turn interactions is limited. In this paper, we investigate whether LLM-based search agents have the ability to communicate their own confidence through verbalized confidence scores after long sequences of actions… ▽ More Confidence in LLMs is a useful indicator of model uncertainty and answer reliability. Existing work mainly focused on single-turn scenarios, while research on confidence in complex multi-turn interactions is limited. In this paper, we investigate whether LLM-based search agents have the ability to communicate their own confidence through verbalized confidence scores after long sequences of actions, a significantly more challenging task compared to outputting confidence in a single interaction. Experimenting on open-source agentic models, we first find that models exhibit much higher task accuracy at high confidence while having near-zero accuracy when confidence is low. Based on this observation, we propose Test-Time Scaling (TTS) methods that use confidence scores to determine answer quality, encourage the model to try again until reaching a satisfactory confidence level. Results show that our proposed methods significantly reduce token consumption while demonstrating competitive performance compared to baseline fixed budget TTS methods. △ Less

Submitted 28 October, 2025; v1 submitted 27 October, 2025; originally announced October 2025.

Comments: 25 pages

E2Rank: Your Text Embedding can Also be an Effective and Efficient Listwise Reranker

Authors: Qi Liu, Yanzhao Zhang, Mingxin Li, Dingkun Long, Pengjun Xie, Jiaxin Mao

Abstract: Text embedding models serve as a fundamental component in real-world search applications. By mapping queries and documents into a shared embedding space, they deliver competitive retrieval performance with high efficiency. However, their ranking fidelity remains limited compared to dedicated rerankers, especially recent LLM-based listwise rerankers, which capture fine-grained query-document and do… ▽ More Text embedding models serve as a fundamental component in real-world search applications. By mapping queries and documents into a shared embedding space, they deliver competitive retrieval performance with high efficiency. However, their ranking fidelity remains limited compared to dedicated rerankers, especially recent LLM-based listwise rerankers, which capture fine-grained query-document and document-document interactions. In this paper, we propose a simple yet effective unified framework E2Rank, means Efficient Embedding-based Ranking (also means Embedding-to-Rank), which extends a single text embedding model to perform both high-quality retrieval and listwise reranking through continued training under a listwise ranking objective, thereby achieving strong effectiveness with remarkable efficiency. By applying cosine similarity between the query and document embeddings as a unified ranking function, the listwise ranking prompt, which is constructed from the original query and its candidate documents, serves as an enhanced query enriched with signals from the top-K documents, akin to pseudo-relevance feedback (PRF) in traditional retrieval models. This design preserves the efficiency and representational quality of the base embedding model while significantly improving its reranking performance. Empirically, E2Rank achieves state-of-the-art results on the BEIR reranking benchmark and demonstrates competitive performance on the reasoning-intensive BRIGHT benchmark, with very low reranking latency. We also show that the ranking training process improves embedding performance on the MTEB benchmark. Our findings indicate that a single embedding model can effectively unify retrieval and reranking, offering both computational efficiency and competitive ranking accuracy. △ Less

Submitted 30 October, 2025; v1 submitted 26 October, 2025; originally announced October 2025.

Comments: Code and models are avaliable at https://alibaba-nlp.github.io/E2Rank

S-Chain: Structured Visual Chain-of-Thought For Medicine

Authors: Khai Le-Duc, Duy M. H. Nguyen, Phuong T. H. Trinh, Tien-Phat Nguyen, Nghiem T. Diep, An Ngo, Tung Vu, Trinh Vuong, Anh-Tien Nguyen, Mau Nguyen, Van Trung Hoang, Khai-Nguyen Nguyen, Hy Nguyen, Chris Ngo, Anji Liu, Nhat Ho, Anne-Christin Hauschild, Khanh Xuan Nguyen, Thanh Nguyen-Tang, Pengtao Xie, Daniel Sonntag, James Zou, Mathias Niepert, Anh Totti Nguyen

Abstract: Faithful reasoning in medical vision-language models (VLMs) requires not only accurate predictions but also transparent alignment between textual rationales and visual evidence. While Chain-of-Thought (CoT) prompting has shown promise in medical visual question answering (VQA), no large-scale expert-level dataset has captured stepwise reasoning with precise visual grounding. We introduce S-Chain,… ▽ More Faithful reasoning in medical vision-language models (VLMs) requires not only accurate predictions but also transparent alignment between textual rationales and visual evidence. While Chain-of-Thought (CoT) prompting has shown promise in medical visual question answering (VQA), no large-scale expert-level dataset has captured stepwise reasoning with precise visual grounding. We introduce S-Chain, the first large-scale dataset of 12,000 expert-annotated medical images with bounding boxes and structured visual CoT (SV-CoT), explicitly linking visual regions to reasoning steps. The dataset further supports 16 languages, totaling over 700k VQA pairs for broad multilingual applicability. Using S-Chain, we benchmark state-of-the-art medical VLMs (ExGra-Med, LLaVA-Med) and general-purpose VLMs (Qwen2.5-VL, InternVL2.5), showing that SV-CoT supervision significantly improves interpretability, grounding fidelity, and robustness. Beyond benchmarking, we study its synergy with retrieval-augmented generation, revealing how domain knowledge and visual grounding interact during autoregressive reasoning. Finally, we propose a new mechanism that strengthens the alignment between visual evidence and reasoning, improving both reliability and efficiency. S-Chain establishes a new benchmark for grounded medical reasoning and paves the way toward more trustworthy and explainable medical VLMs. △ Less

Submitted 26 October, 2025; originally announced October 2025.

Comments: First version

DecoupleSearch: Decouple Planning and Search via Hierarchical Reward Modeling

Authors: Hao Sun, Zile Qiao, Bo Wang, Guoxin Chen, Yingyan Hou, Yong Jiang, Pengjun Xie, Fei Huang, Yan Zhang

Abstract: Retrieval-Augmented Generation (RAG) systems have emerged as a pivotal methodology for enhancing Large Language Models (LLMs) through the dynamic integration of external knowledge. To further improve RAG's flexibility, Agentic RAG introduces autonomous agents into the workflow. However, Agentic RAG faces several challenges: (1) the success of each step depends on both high-quality planning and acc… ▽ More Retrieval-Augmented Generation (RAG) systems have emerged as a pivotal methodology for enhancing Large Language Models (LLMs) through the dynamic integration of external knowledge. To further improve RAG's flexibility, Agentic RAG introduces autonomous agents into the workflow. However, Agentic RAG faces several challenges: (1) the success of each step depends on both high-quality planning and accurate search, (2) the lack of supervision for intermediate reasoning steps, and (3) the exponentially large candidate space for planning and searching. To address these challenges, we propose DecoupleSearch, a novel framework that decouples planning and search processes using dual value models, enabling independent optimization of plan reasoning and search grounding. Our approach constructs a reasoning tree, where each node represents planning and search steps. We leverage Monte Carlo Tree Search to assess the quality of each step. During inference, Hierarchical Beam Search iteratively refines planning and search candidates with dual value models. Extensive experiments across policy models of varying parameter sizes, demonstrate the effectiveness of our method. △ Less

Submitted 7 September, 2025; originally announced October 2025.

Comments: EMNLP 2025 Main Conference

PIRA: Pan-CDN Intra-video Resource Adaptation for Short Video Streaming

Authors: Chunyu Qiao, Tong Liu, Yucheng Zhang, Zhiwei Fan, Pengjin Xie, Zhen Wang, Liang Liu

Abstract: In large scale short video platforms, CDN resource selection plays a critical role in maintaining Quality of Experience (QoE) while controlling escalating traffic costs. To better understand this phenomenon, we conduct in the wild network measurements during video playback in a production short video system. The results reveal that CDNs delivering higher average QoE often come at greater financial… ▽ More In large scale short video platforms, CDN resource selection plays a critical role in maintaining Quality of Experience (QoE) while controlling escalating traffic costs. To better understand this phenomenon, we conduct in the wild network measurements during video playback in a production short video system. The results reveal that CDNs delivering higher average QoE often come at greater financial cost, yet their connection quality fluctuates even within a single video underscoring a fundamental and dynamic trade off between QoE and cost. However, the problem of sustaining high QoE under cost constraints remains insufficiently investigated in the context of CDN selection for short video streaming. To address this, we propose PIRA, a dynamic resource selection algorithm that optimizes QoE and cost in real time during video playback. PIRA formally integrating QoE and cost by a mathematical model, and introduce a intra video control theoretic CDN resource selection approach which can balance QoE and cost under network dynamics. To reduce the computation overheads, PIRA employs state space pruning and adaptive parameter adjustment to efficiently solve the high dimensional optimization problem. In large scale production experiments involving 450,000 users over two weeks, PIRA outperforms the production baseline, achieving a 2.1% reduction in start up delay, 15.2% shorter rebuffering time, and 10% lower average unit traffic cost, demonstrating its effectiveness in balancing user experience and financial cost at scale. △ Less

Submitted 21 October, 2025; originally announced October 2025.

DeLoad: Demand-Driven Short-Video Preloading with Scalable Watch-Time Estimation

Authors: Tong Liu, Zhiwei Fan, Guanyan Peng, Haodan Zhang, Yucheng Zhang, Zhen Wang, Pengjin Xie, Liang Liu

Abstract: Short video streaming has become a dominant paradigm in digital media, characterized by rapid swiping interactions and diverse media content. A key technical challenge is designing an effective preloading strategy that dynamically selects and prioritizes download tasks from an evolving playlist, balancing Quality of Experience (QoE) and bandwidth efficiency under practical commercial constraints.… ▽ More Short video streaming has become a dominant paradigm in digital media, characterized by rapid swiping interactions and diverse media content. A key technical challenge is designing an effective preloading strategy that dynamically selects and prioritizes download tasks from an evolving playlist, balancing Quality of Experience (QoE) and bandwidth efficiency under practical commercial constraints. However, real world analysis reveals critical limitations of existing approaches: (1) insufficient adaptation of download task sizes to dynamic conditions, and (2) watch time prediction models that are difficult to deploy reliably at scale. In this paper, we propose DeLoad, a novel preloading framework that addresses these issues by introducing dynamic task sizing and a practical, multi dimensional watch time estimation method. Additionally, a Deep Reinforcement Learning (DRL) enhanced agent is trained to optimize the download range decisions adaptively. Extensive evaluations conducted on an offline testing platform, leveraging massive real world network data, demonstrate that DeLoad achieves significant improvements in QoE metrics (34.4% to 87.4% gain). Furthermore, after deployment on a large scale commercial short video platform, DeLoad has increased overall user watch time by 0.09% while simultaneously reducing rebuffering events and 3.76% bandwidth consumption. △ Less

Submitted 21 October, 2025; originally announced October 2025.

Supervised Fine-Tuning or Contrastive Learning? Towards Better Multimodal LLM Reranking

Authors: Ziqi Dai, Xin Zhang, Mingxin Li, Yanzhao Zhang, Dingkun Long, Pengjun Xie, Meishan Zhang, Wenjie Li, Min Zhang

Abstract: In information retrieval, training reranking models mainly focuses on two types of objectives: metric learning (e.g. contrastive loss to increase the predicted scores on relevant query-document pairs) and classification (binary label prediction of relevance vs. irrelevance). For BERT-style encoders, various studies have shown that contrastive learning (CL) can be more effective than discriminative… ▽ More In information retrieval, training reranking models mainly focuses on two types of objectives: metric learning (e.g. contrastive loss to increase the predicted scores on relevant query-document pairs) and classification (binary label prediction of relevance vs. irrelevance). For BERT-style encoders, various studies have shown that contrastive learning (CL) can be more effective than discriminative (classification) learning. However, for large language models (LLMs), classification via supervised fine-tuning (SFT), which predicts ''yes'' (resp. ''no'') token for relevant (resp. irrelevant) pairs, appears more promising as it aligns well with the generative nature of LLMs. This divergence raises a central question: which objective is intrinsically better suited to LLM-based reranking, and what mechanism underlies the difference? In this work, we conduct a comprehensive comparison and analysis between CL and SFT for reranking, taking the universal multimodal retrieval (UMR) as the experimental playground. We first decompose the objectives into two components: weight, which controls the magnitude of those updates, and direction, which guides the model updates, then present a unified framework for understanding their interactions. Through probing experiments, we find that SFT provides a substantially stronger weighting scheme than CL, whereas the preferred scoring direction shows no clear winner. Taken together, these results point to a consistent advantage of SFT over CL for LLM reranking. To further validate our findings, we conduct large-scale training with SFT and present new state-of-the-art rerankers on the MRB benchmark. We also provide ablations on SFT settings and expect our findings to benefit future research and applications in this area. △ Less

Submitted 16 October, 2025; originally announced October 2025.

Qwen3Guard Technical Report

Authors: Haiquan Zhao, Chenhan Yuan, Fei Huang, Xiaomeng Hu, Yichang Zhang, An Yang, Bowen Yu, Dayiheng Liu, Jingren Zhou, Junyang Lin, Baosong Yang, Chen Cheng, Jialong Tang, Jiandong Jiang, Jianwei Zhang, Jijie Xu, Ming Yan, Minmin Sun, Pei Zhang, Pengjun Xie, Qiaoyu Tang, Qin Zhu, Rong Zhang, Shibin Wu, Shuo Zhang , et al. (18 additional authors not shown)

Abstract: As large language models (LLMs) become more capable and widely used, ensuring the safety of their outputs is increasingly critical. Existing guardrail models, though useful in static evaluation settings, face two major limitations in real-world applications: (1) they typically output only binary "safe/unsafe" labels, which can be interpreted inconsistently across diverse safety policies, rendering… ▽ More As large language models (LLMs) become more capable and widely used, ensuring the safety of their outputs is increasingly critical. Existing guardrail models, though useful in static evaluation settings, face two major limitations in real-world applications: (1) they typically output only binary "safe/unsafe" labels, which can be interpreted inconsistently across diverse safety policies, rendering them incapable of accommodating varying safety tolerances across domains; and (2) they require complete model outputs before performing safety checks, making them fundamentally incompatible with streaming LLM inference, thereby preventing timely intervention during generation and increasing exposure to harmful partial outputs. To address these challenges, we present Qwen3Guard, a series of multilingual safety guardrail models with two specialized variants: Generative Qwen3Guard, which casts safety classification as an instruction-following task to enable fine-grained tri-class judgments (safe, controversial, unsafe); and Stream Qwen3Guard, which introduces a token-level classification head for real-time safety monitoring during incremental text generation. Both variants are available in three sizes (0.6B, 4B, and 8B parameters) and support up to 119 languages and dialects, providing comprehensive, scalable, and low-latency safety moderation for global LLM deployments. Evaluated across English, Chinese, and multilingual benchmarks, Qwen3Guard achieves state-of-the-art performance in both prompt and response safety classification. All models are released under the Apache 2.0 license for public use. △ Less

Submitted 16 October, 2025; originally announced October 2025.

More than A Point: Capturing Uncertainty with Adaptive Affordance Heatmaps for Spatial Grounding in Robotic Tasks

Authors: Xinyu Shao, Yanzhe Tang, Pengwei Xie, Kaiwen Zhou, Yuzheng Zhuang, Xingyue Quan, Jianye Hao, Long Zeng, Xiu Li

Abstract: Many language-guided robotic systems rely on collapsing spatial reasoning into discrete points, making them brittle to perceptual noise and semantic ambiguity. To address this challenge, we propose RoboMAP, a framework that represents spatial targets as continuous, adaptive affordance heatmaps. This dense representation captures the uncertainty in spatial grounding and provides richer information… ▽ More Many language-guided robotic systems rely on collapsing spatial reasoning into discrete points, making them brittle to perceptual noise and semantic ambiguity. To address this challenge, we propose RoboMAP, a framework that represents spatial targets as continuous, adaptive affordance heatmaps. This dense representation captures the uncertainty in spatial grounding and provides richer information for downstream policies, thereby significantly enhancing task success and interpretability. RoboMAP surpasses the previous state-of-the-art on a majority of grounding benchmarks with up to a 50x speed improvement, and achieves an 82\% success rate in real-world manipulation. Across extensive simulated and physical experiments, it demonstrates robust performance and shows strong zero-shot generalization to navigation. More details and videos can be found at https://robo-map.github.io. △ Less

Submitted 15 October, 2025; v1 submitted 12 October, 2025; originally announced October 2025.

Comments: More details and videos can be found at https://robo-map.github.io

RepDL: Bit-level Reproducible Deep Learning Training and Inference

Authors: Peichen Xie, Xian Zhang, Shuo Chen

Abstract: Non-determinism and non-reproducibility present significant challenges in deep learning, leading to inconsistent results across runs and platforms. These issues stem from two origins: random number generation and floating-point computation. While randomness can be controlled through deterministic configurations, floating-point inconsistencies remain largely unresolved. To address this, we introduc… ▽ More Non-determinism and non-reproducibility present significant challenges in deep learning, leading to inconsistent results across runs and platforms. These issues stem from two origins: random number generation and floating-point computation. While randomness can be controlled through deterministic configurations, floating-point inconsistencies remain largely unresolved. To address this, we introduce RepDL, an open-source library that ensures deterministic and bitwise-reproducible deep learning training and inference across diverse computing environments. RepDL achieves this by enforcing correct rounding and order invariance in floating-point computation. The source code is available at https://github.com/microsoft/RepDL . △ Less

Submitted 10 October, 2025; originally announced October 2025.

Comments: Originally drafted in 2023

Demystifying deep search: a holistic evaluation with hint-free multi-hop questions and factorised metrics

Authors: Maojia Song, Renhang Liu, Xinyu Wang, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou, Dorien Herremans, Soujanya Poria

Abstract: RAG (Retrieval-Augmented Generation) systems and web agents are increasingly evaluated on multi-hop deep search tasks, yet current practice suffers from two major limitations. First, most benchmarks leak the reasoning path in the question text, allowing models to follow surface cues rather than discover reasoning chains autonomously. Second, evaluation is typically reduced to a single pass rate, w… ▽ More RAG (Retrieval-Augmented Generation) systems and web agents are increasingly evaluated on multi-hop deep search tasks, yet current practice suffers from two major limitations. First, most benchmarks leak the reasoning path in the question text, allowing models to follow surface cues rather than discover reasoning chains autonomously. Second, evaluation is typically reduced to a single pass rate, which collapses diverse behaviours into one score and obscures whether failures stem from inadequate search, poor knowledge use, or inappropriate refusal. To address these issues, we present WebDetective, a benchmark of hint-free multi-hop questions paired with a controlled Wikipedia sandbox that ensures full traceability of model actions, and a holistic evaluation framework that separates search sufficiency, knowledge utilisation, and refusal behaviour. Our evaluation of 25 state-of-the-art models reveals systematic weaknesses across all architectures: models struggle with knowledge utilisation despite having sufficient evidence and demonstrate near-absent appropriate refusal when evidence is lacking. These patterns expose a fundamental gap: today's systems excel at executing given reasoning paths but fail when required to discover them. We develop an agentic workflow, EvidenceLoop, that explicitly targets the challenges our benchmark identifies, incorporating verification loops and systematic evidence tracking that improve both search and synthesis capabilities. This baseline demonstrates that WebDetective's diagnostic framework can guide concrete architectural improvements, establishing our benchmark as a critical tool for developing genuinely autonomous reasoning systems rather than pattern-following agents. △ Less

Submitted 10 December, 2025; v1 submitted 1 October, 2025; originally announced October 2025.

MARS: Optimizing Dual-System Deep Research via Multi-Agent Reinforcement Learning

Authors: Guoxin Chen, Zile Qiao, Wenqing Wang, Donglei Yu, Xuanzhong Chen, Hao Sun, Minpeng Liao, Kai Fan, Yong Jiang, Penguin Xie, Wayne Xin Zhao, Ruihua Song, Fei Huang

Abstract: Large Reasoning Models (LRMs) often exhibit a tendency for overanalysis in simple tasks, where the models excessively utilize System 2-type, deliberate reasoning, leading to inefficient token generation. Furthermore, these models face challenges in adapting their reasoning capabilities to rapidly changing environments due to the static nature of their pretraining data. To address these issues, adv… ▽ More Large Reasoning Models (LRMs) often exhibit a tendency for overanalysis in simple tasks, where the models excessively utilize System 2-type, deliberate reasoning, leading to inefficient token generation. Furthermore, these models face challenges in adapting their reasoning capabilities to rapidly changing environments due to the static nature of their pretraining data. To address these issues, advancing Large Language Models (LLMs) for complex reasoning tasks requires innovative approaches that bridge intuitive and deliberate cognitive processes, akin to human cognition's dual-system dynamic. This paper introduces a Multi-Agent System for Deep ReSearch (MARS) enabling seamless integration of System 1's fast, intuitive thinking with System 2's deliberate reasoning within LLMs. MARS strategically integrates multiple external tools, such as Google Search, Google Scholar, and Python Interpreter, to access up-to-date information and execute complex computations, while creating a specialized division of labor where System 1 efficiently processes and summarizes high-volume external information, providing distilled insights that expand System 2's reasoning context without overwhelming its capacity. Furthermore, we propose a multi-agent reinforcement learning framework extending Group Relative Policy Optimization to simultaneously optimize both systems with multi-turn tool interactions, bin-packing optimization, and sample balancing strategies that enhance collaborative efficiency. Extensive experiments demonstrate MARS achieves substantial improvements of 3.86% on the challenging Humanity's Last Exam (HLE) benchmark and an average gain of 8.9% across 7 knowledge-intensive tasks, validating the effectiveness of our dual-system paradigm for complex reasoning in dynamic information environments. △ Less

Submitted 6 October, 2025; originally announced October 2025.

Comments: Ongoing Work

Can Prompts Rewind Time for LLMs? Evaluating the Effectiveness of Prompted Knowledge Cutoffs

Authors: Xin Gao, Ruiyi Zhang, Daniel Du, Saurabh Mahindre, Sai Ashish Somayajula, Pengtao Xie

Abstract: Large Language Models (LLMs) are widely used for temporal prediction, but their reliance on pretraining data raises contamination concerns, as accurate predictions on pre-cutoff test data may reflect memorization rather than reasoning, leading to an overestimation of their generalization capability. With the recent emergence of prompting-based unlearning techniques, a natural question arises: Can… ▽ More Large Language Models (LLMs) are widely used for temporal prediction, but their reliance on pretraining data raises contamination concerns, as accurate predictions on pre-cutoff test data may reflect memorization rather than reasoning, leading to an overestimation of their generalization capability. With the recent emergence of prompting-based unlearning techniques, a natural question arises: Can LLMs be prompted to simulate an earlier knowledge cutoff? In this work, we investigate the capability of prompting to simulate earlier knowledge cutoff in LLMs. We construct three evaluation datasets to assess the extent to which LLMs can forget (1) direct factual knowledge, (2) semantic shifts, and (3) causally related knowledge. Results demonstrate that while prompt-based simulated knowledge cutoffs show effectiveness when directly queried with the information after that date, they struggle to induce forgetting when the forgotten content is not directly asked but causally related to the query. These findings highlight the need for more rigorous evaluation settings when applying LLMs for temporal prediction tasks. The full dataset and evaluation code are available at https://github.com/gxx27/time_unlearn. △ Less

Submitted 14 October, 2025; v1 submitted 26 September, 2025; originally announced October 2025.

Comments: Published at EMNLP 2025; Code and data available at https://github.com/gxx27/time_unlearn

Scaling Generalist Data-Analytic Agents

Authors: Shuofei Qiao, Yanqiu Zhao, Zhisong Qiu, Xiaobin Wang, Jintian Zhang, Zhao Bin, Ningyu Zhang, Yong Jiang, Pengjun Xie, Fei Huang, Huajun Chen

Abstract: Data-analytic agents are emerging as a key catalyst for automated scientific discovery and for the vision of Innovating AI. Current approaches, however, rely heavily on prompt engineering over proprietary models, while open-source models struggle to face diverse-format, large-scale data files and long-horizon, multi-step reasoning that real-world analytics demands. This paper introduces DataMind,… ▽ More Data-analytic agents are emerging as a key catalyst for automated scientific discovery and for the vision of Innovating AI. Current approaches, however, rely heavily on prompt engineering over proprietary models, while open-source models struggle to face diverse-format, large-scale data files and long-horizon, multi-step reasoning that real-world analytics demands. This paper introduces DataMind, a scalable data synthesis and agent training recipe designed to build generalist data-analytic agents. DataMind tackles three key challenges in building open-source data-analytic agents, including insufficient data resources, improper training strategy, and unstable code-based multi-turn rollout. Concretely, DataMind applies 1) a fine-grained task taxonomy and a recursive easy-to-hard task composition mechanism to increase the diversity and difficulty of synthesized queries; 2) a knowledge-augmented trajectory sampling strategy followed by model-based and rule-based filtering; 3) a dynamically adjustable training objective combining both SFT and RL losses; 4) a memory-frugal and stable code-based multi-turn rollout framework. Built on DataMind, we curate DataMind-12K, a high-quality trajectory set spanning diverse domains, task categories, and data file formats for data-analytic tasks. Trained on DataMind-12K, our DataMind-14B achieves state-of-the-art with an average score of 71.16% on multiple data analysis benchmarks, outperforming the strongest proprietary baselines DeepSeek-V3.1 and GPT-5. Our DataMind-7B also performs best among all open-source models with a score of 68.10%. We also incorporate some empirical insights gained from our exploratory trials into the analysis experiments, aiming to provide actionable insights about agentic training for the community. We will release DataMind-12K and DataMind-7B,14B for the community's future research. △ Less

Submitted 29 September, 2025; originally announced September 2025.

Comments: Work in progress

ReSum: Unlocking Long-Horizon Search Intelligence via Context Summarization

Authors: Xixi Wu, Kuan Li, Yida Zhao, Liwen Zhang, Litu Ou, Huifeng Yin, Zhongwang Zhang, Xinmiao Yu, Dingchu Zhang, Yong Jiang, Pengjun Xie, Fei Huang, Minhao Cheng, Shuai Wang, Hong Cheng, Jingren Zhou

Abstract: Large Language Model (LLM)-based web agents demonstrate strong performance on knowledge-intensive tasks but are hindered by context window limitations in paradigms like ReAct. Complex queries involving multiple entities, intertwined relationships, and high uncertainty demand extensive search cycles that rapidly exhaust context budgets before reaching solutions. To overcome this challenge, we intro… ▽ More Large Language Model (LLM)-based web agents demonstrate strong performance on knowledge-intensive tasks but are hindered by context window limitations in paradigms like ReAct. Complex queries involving multiple entities, intertwined relationships, and high uncertainty demand extensive search cycles that rapidly exhaust context budgets before reaching solutions. To overcome this challenge, we introduce ReSum, a novel paradigm that enables indefinite exploration through periodic context summarization. ReSum converts growing interaction histories into compact reasoning states, maintaining awareness of prior discoveries while bypassing context constraints. For paradigm adaptation, we propose ReSum-GRPO, integrating GRPO with segmented trajectory training and advantage broadcasting to familiarize agents with summary-conditioned reasoning. Extensive experiments on web agents across three benchmarks demonstrate that ReSum delivers an average absolute improvement of 4.5% over ReAct, with further gains of 8.2% following ReSum-GRPO training. Notably, with only 1K training samples, our WebResummer-30B (a ReSum-GRPO-trained version of WebSailor-30B) achieves 33.3% Pass@1 on BrowseComp-zh and 18.3% on BrowseComp-en, surpassing most open-source web agents. △ Less

Submitted 15 October, 2025; v1 submitted 16 September, 2025; originally announced September 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

WebWeaver: Structuring Web-Scale Evidence with Dynamic Outlines for Open-Ended Deep Research

Authors: Zijian Li, Xin Guan, Bo Zhang, Shen Huang, Houquan Zhou, Shaopeng Lai, Ming Yan, Yong Jiang, Pengjun Xie, Fei Huang, Jun Zhang, Jingren Zhou

Abstract: This paper tackles \textbf{open-ended deep research (OEDR)}, a complex challenge where AI agents must synthesize vast web-scale information into insightful reports. Current approaches are plagued by dual-fold limitations: static research pipelines that decouple planning from evidence acquisition and monolithic generation paradigms that include redundant, irrelevant evidence, suffering from halluci… ▽ More This paper tackles \textbf{open-ended deep research (OEDR)}, a complex challenge where AI agents must synthesize vast web-scale information into insightful reports. Current approaches are plagued by dual-fold limitations: static research pipelines that decouple planning from evidence acquisition and monolithic generation paradigms that include redundant, irrelevant evidence, suffering from hallucination issues and low citation accuracy. To address these challenges, we introduce \textbf{WebWeaver}, a novel dual-agent framework that emulates the human research process. The planner operates in a dynamic cycle, iteratively interleaving evidence acquisition with outline optimization to produce a comprehensive, citation-grounded outline linking to a memory bank of evidence. The writer then executes a hierarchical retrieval and writing process, composing the report section by section. By performing targeted retrieval of only the necessary evidence from the memory bank via citations for each part, it effectively mitigates long-context issues and citation hallucinations. Our framework establishes a new state-of-the-art across major OEDR benchmarks, including DeepResearch Bench, DeepConsult, and DeepResearchGym. These results validate our human-centric, iterative methodology, demonstrating that adaptive planning and focused synthesis are crucial for producing comprehensive, trusted, and well-structured reports. △ Less

Submitted 7 October, 2025; v1 submitted 16 September, 2025; originally announced September 2025.

Comments: An agent system for open-ended deep research

Towards General Agentic Intelligence via Environment Scaling

Authors: Runnan Fang, Shihao Cai, Baixuan Li, Jialong Wu, Guangyu Li, Wenbiao Yin, Xinyu Wang, Xiaobin Wang, Liangcai Su, Zhen Zhang, Shibin Wu, Zhengwei Tao, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Advanced agentic intelligence is a prerequisite for deploying Large Language Models in practical, real-world applications. Diverse real-world APIs demand precise, robust function-calling intelligence, which needs agents to develop these capabilities through interaction in varied environments. The breadth of function-calling competence is closely tied to the diversity of environments in which agent… ▽ More Advanced agentic intelligence is a prerequisite for deploying Large Language Models in practical, real-world applications. Diverse real-world APIs demand precise, robust function-calling intelligence, which needs agents to develop these capabilities through interaction in varied environments. The breadth of function-calling competence is closely tied to the diversity of environments in which agents are trained. In this work, we scale up environments as a step towards advancing general agentic intelligence. This gives rise to two central challenges: (i) how to scale environments in a principled manner, and (ii) how to effectively train agentic capabilities from experiences derived through interactions with these environments. To address these, we design a scalable framework that automatically constructs heterogeneous environments that are fully simulated, systematically broadening the space of function-calling scenarios. We further adapt a two-phase agent fine-tuning strategy: first endowing agents with fundamental agentic capabilities, then specializing them for domain-specific contexts. Extensive experiments on agentic benchmarks, tau-bench, tau2-Bench, and ACEBench, demonstrate that our trained model, AgentScaler, significantly enhances the function-calling capability of models. △ Less

Submitted 16 September, 2025; originally announced September 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

Scaling Agents via Continual Pre-training

Authors: Liangcai Su, Zhen Zhang, Guangyu Li, Zhuo Chen, Chenxi Wang, Maojia Song, Xinyu Wang, Kuan Li, Jialong Wu, Xuanzhong Chen, Zile Qiao, Zhongwang Zhang, Huifeng Yin, Shihao Cai, Runnan Fang, Zhengwei Tao, Wenbiao Yin, Chenxiong Qian, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Large language models (LLMs) have evolved into agentic systems capable of autonomous tool use and multi-step reasoning for complex problem-solving. However, post-training approaches building upon general-purpose foundation models consistently underperform in agentic tasks, particularly in open-source implementations. We identify the root cause: the absence of robust agentic foundation models force… ▽ More Large language models (LLMs) have evolved into agentic systems capable of autonomous tool use and multi-step reasoning for complex problem-solving. However, post-training approaches building upon general-purpose foundation models consistently underperform in agentic tasks, particularly in open-source implementations. We identify the root cause: the absence of robust agentic foundation models forces models during post-training to simultaneously learn diverse agentic behaviors while aligning them to expert demonstrations, thereby creating fundamental optimization tensions. To this end, we are the first to propose incorporating Agentic Continual Pre-training (Agentic CPT) into the deep research agents training pipeline to build powerful agentic foundational models. Based on this approach, we develop a deep research agent model named AgentFounder. We evaluate our AgentFounder-30B on 10 benchmarks and achieve state-of-the-art performance while retains strong tool-use ability, notably 39.9% on BrowseComp-en, 43.3% on BrowseComp-zh, and 31.5% Pass@1 on HLE. △ Less

Submitted 16 September, 2025; originally announced September 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

WebResearcher: Unleashing unbounded reasoning capability in Long-Horizon Agents

Authors: Zile Qiao, Guoxin Chen, Xuanzhong Chen, Donglei Yu, Wenbiao Yin, Xinyu Wang, Zhen Zhang, Baixuan Li, Huifeng Yin, Kuan Li, Rui Min, Minpeng Liao, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Recent advances in deep-research systems have demonstrated the potential for AI agents to autonomously discover and synthesize knowledge from external sources. In this paper, we introduce WebResearcher, a novel framework for building such agents through two key components: (1) WebResearcher, an iterative deep-research paradigm that reformulates deep research as a Markov Decision Process, where age… ▽ More Recent advances in deep-research systems have demonstrated the potential for AI agents to autonomously discover and synthesize knowledge from external sources. In this paper, we introduce WebResearcher, a novel framework for building such agents through two key components: (1) WebResearcher, an iterative deep-research paradigm that reformulates deep research as a Markov Decision Process, where agents periodically consolidate findings into evolving reports while maintaining focused workspaces, overcoming the context suffocation and noise contamination that plague existing mono-contextual approaches; and (2) WebFrontier, a scalable data synthesis engine that generates high-quality training data through tool-augmented complexity escalation, enabling systematic creation of research tasks that bridge the gap between passive knowledge recall and active knowledge construction. Notably, we find that the training data from our paradigm significantly enhances tool-use capabilities even for traditional mono-contextual methods. Furthermore, our paradigm naturally scales through parallel thinking, enabling concurrent multi-agent exploration for more comprehensive conclusions. Extensive experiments across 6 challenging benchmarks demonstrate that WebResearcher achieves state-of-the-art performance, even surpassing frontier proprietary systems. △ Less

Submitted 20 September, 2025; v1 submitted 16 September, 2025; originally announced September 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

WebSailor-V2: Bridging the Chasm to Proprietary Agents via Synthetic Data and Scalable Reinforcement Learning

Authors: Kuan Li, Zhongwang Zhang, Huifeng Yin, Rui Ye, Yida Zhao, Liwen Zhang, Litu Ou, Dingchu Zhang, Xixi Wu, Jialong Wu, Xinyu Wang, Zile Qiao, Zhen Zhang, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Transcending human cognitive limitations represents a critical frontier in LLM training. Proprietary agentic systems like DeepResearch have demonstrated superhuman capabilities on extremely complex information-seeking benchmarks such as BrowseComp, a feat previously unattainable. We posit that their success hinges on a sophisticated reasoning pattern absent in open-source models: the ability to sy… ▽ More Transcending human cognitive limitations represents a critical frontier in LLM training. Proprietary agentic systems like DeepResearch have demonstrated superhuman capabilities on extremely complex information-seeking benchmarks such as BrowseComp, a feat previously unattainable. We posit that their success hinges on a sophisticated reasoning pattern absent in open-source models: the ability to systematically reduce extreme uncertainty when navigating vast information landscapes. Based on this insight, we introduce WebSailor, a complete post-training methodology designed to instill this crucial capability. Our approach involves generating novel, high-uncertainty tasks through structured sampling and information obfuscation, RFT cold start, and an efficient agentic RL training algorithm, Duplicating Sampling Policy Optimization (DUPO). With this integrated pipeline, WebSailor significantly outperforms all open-source agents in complex information-seeking tasks, matching proprietary agents' performance and closing the capability gap. △ Less

Submitted 16 September, 2025; originally announced September 2025.

Comments: https://tongyi-agent.github.io/blog/introducing-tongyi-deep-research/

OmniEVA: Embodied Versatile Planner via Task-Adaptive 3D-Grounded and Embodiment-aware Reasoning

Authors: Yuecheng Liu, Dafeng Chi, Shiguang Wu, Zhanguang Zhang, Yuzheng Zhuang, Bowen Yang, He Zhu, Lingfeng Zhang, Pengwei Xie, David Gamaliel Arcos Bravo, Yingxue Zhang, Jianye Hao, Xingyue Quan

Abstract: Recent advances in multimodal large language models (MLLMs) have opened new opportunities for embodied intelligence, enabling multimodal understanding, reasoning, and interaction, as well as continuous spatial decision-making. Nevertheless, current MLLM-based embodied systems face two critical limitations. First, Geometric Adaptability Gap: models trained solely on 2D inputs or with hard-coded 3D… ▽ More Recent advances in multimodal large language models (MLLMs) have opened new opportunities for embodied intelligence, enabling multimodal understanding, reasoning, and interaction, as well as continuous spatial decision-making. Nevertheless, current MLLM-based embodied systems face two critical limitations. First, Geometric Adaptability Gap: models trained solely on 2D inputs or with hard-coded 3D geometry injection suffer from either insufficient spatial information or restricted 2D generalization, leading to poor adaptability across tasks with diverse spatial demands. Second, Embodiment Constraint Gap: prior work often neglects the physical constraints and capacities of real robots, resulting in task plans that are theoretically valid but practically infeasible. To address these gaps, we introduce OmniEVA -- an embodied versatile planner that enables advanced embodied reasoning and task planning through two pivotal innovations: (1) a Task-Adaptive 3D Grounding mechanism, which introduces a gated router to perform explicit selective regulation of 3D fusion based on contextual requirements, enabling context-aware 3D grounding for diverse embodied tasks. (2) an Embodiment-Aware Reasoning framework that jointly incorporates task goals and embodiment constraints into the reasoning loop, resulting in planning decisions that are both goal-directed and executable. Extensive experimental results demonstrate that OmniEVA not only achieves state-of-the-art general embodied reasoning performance, but also exhibits a strong ability across a wide range of downstream scenarios. Evaluations of a suite of proposed embodied benchmarks, including both primitive and composite tasks, confirm its robust and versatile planning capabilities. Project page: https://omnieva.github.io △ Less

Submitted 12 September, 2025; v1 submitted 11 September, 2025; originally announced September 2025.

TextlessRAG: End-to-End Visual Document RAG by Speech Without Text

Authors: Peijin Xie, Shun Qian, Bingquan Liu, Dexin Wang, Lin Sun, Xiangzheng Zhang

Abstract: Document images encapsulate a wealth of knowledge, while the portability of spoken queries enables broader and flexible application scenarios. Yet, no prior work has explored knowledge base question answering over visual document images with queries provided directly in speech. We propose TextlessRAG, the first end-to-end framework for speech-based question answering over large-scale document imag… ▽ More Document images encapsulate a wealth of knowledge, while the portability of spoken queries enables broader and flexible application scenarios. Yet, no prior work has explored knowledge base question answering over visual document images with queries provided directly in speech. We propose TextlessRAG, the first end-to-end framework for speech-based question answering over large-scale document images. Unlike prior methods, TextlessRAG eliminates ASR, TTS and OCR, directly interpreting speech, retrieving relevant visual knowledge, and generating answers in a fully textless pipeline. To further boost performance, we integrate a layout-aware reranking mechanism to refine retrieval. Experiments demonstrate substantial improvements in both efficiency and accuracy. To advance research in this direction, we also release the first bilingual speech--document RAG dataset, featuring Chinese and English voice queries paired with multimodal document content. Both the dataset and our pipeline will be made available at repository:https://github.com/xiepeijinhit-hue/textlessrag △ Less

Submitted 10 September, 2025; v1 submitted 9 September, 2025; originally announced September 2025.

Comments: 5 pages, 4 figures,

Robust Docking Maneuvers for Autonomous Trolley Collection: An Optimization-Based Visual Servoing Scheme

Authors: Yuhan Pang, Bingyi Xia, Zhe Zhang, Zhirui Sun, Peijia Xie, Bike Zhu, Wenjun Xu, Jiankun Wang

Abstract: Service robots have demonstrated significant potential for autonomous trolley collection and redistribution in public spaces like airports or warehouses to improve efficiency and reduce cost. Usually, a fully autonomous system for the collection and transportation of multiple trolleys is based on a Leader-Follower formation of mobile manipulators, where reliable docking maneuvers of the mobile bas… ▽ More Service robots have demonstrated significant potential for autonomous trolley collection and redistribution in public spaces like airports or warehouses to improve efficiency and reduce cost. Usually, a fully autonomous system for the collection and transportation of multiple trolleys is based on a Leader-Follower formation of mobile manipulators, where reliable docking maneuvers of the mobile base are essential to align trolleys into organized queues. However, developing a vision-based robotic docking system faces significant challenges: high precision requirements, environmental disturbances, and inherent robot constraints. To address these challenges, we propose an optimization-based Visual Servoing scheme that incorporates active infrared markers for robust feature extraction across diverse lighting conditions. This framework explicitly models nonholonomic kinematics and visibility constraints within the Hybrid Visual Servoing problem, augmented with an observer for disturbance rejection to ensure precise and stable docking. Experimental results across diverse environments demonstrate the robustness of this system, with quantitative evaluations confirming high docking accuracy. △ Less

Submitted 17 September, 2025; v1 submitted 9 September, 2025; originally announced September 2025.

Domain-Aware RAG: MoL-Enhanced RL for Efficient Training and Scalable Retrieval

Authors: Hao Lin, Peitong Xie, Jingxue Chen, Jie Lin, Qingkun Tang, Qianchun Lu

Abstract: Retrieval-Augmented Generation (RAG) systems rely heavily on the retrieval stage, particularly the coarse-ranking process. Existing coarse-ranking optimization approaches often struggle to balance domain-specific knowledge learning with query enhencement, resulting in suboptimal retrieval performance. To address this challenge, we propose MoLER, a domain-aware RAG method that uses MoL-Enhanced Rei… ▽ More Retrieval-Augmented Generation (RAG) systems rely heavily on the retrieval stage, particularly the coarse-ranking process. Existing coarse-ranking optimization approaches often struggle to balance domain-specific knowledge learning with query enhencement, resulting in suboptimal retrieval performance. To address this challenge, we propose MoLER, a domain-aware RAG method that uses MoL-Enhanced Reinforcement Learning to optimize retrieval. MoLER has a two-stage pipeline: a continual pre-training (CPT) phase using a Mixture of Losses (MoL) to balance domain-specific knowledge with general language capabilities, and a reinforcement learning (RL) phase leveraging Group Relative Policy Optimization (GRPO) to optimize query and passage generation for maximizing document recall. A key innovation is our Multi-query Single-passage Late Fusion (MSLF) strategy, which reduces computational overhead during RL training while maintaining scalable inference via Multi-query Multi-passage Late Fusion (MMLF). Extensive experiments on benchmark datasets show that MoLER achieves state-of-the-art performance, significantly outperforming baseline methods. MoLER bridges the knowledge gap in RAG systems, enabling robust and scalable retrieval in specialized domains. △ Less

Submitted 8 September, 2025; originally announced September 2025.

DreamPRM-1.5: Unlocking the Potential of Each Instance for Multimodal Process Reward Model Training

Authors: Qi Cao, Pengtao Xie

Abstract: Training multimodal process reward models (PRMs) is hard due to (i) distribution shift between training set and test set and (ii) quality imbalance across training data samples. While domain-level reweighting (e.g., DreamPRM) aligns training with test-time objectives, it leaves a clear gap to an oracle upper bound (pass@N), even under a "sanity check" that uses test set data to probe headroom -- p… ▽ More Training multimodal process reward models (PRMs) is hard due to (i) distribution shift between training set and test set and (ii) quality imbalance across training data samples. While domain-level reweighting (e.g., DreamPRM) aligns training with test-time objectives, it leaves a clear gap to an oracle upper bound (pass@N), even under a "sanity check" that uses test set data to probe headroom -- pointing to meta-level under-parameterization. We introduce DreamPRM-1.5, an instance-level reweighting framework that assigns an adaptive weight to every training example via bi-level optimization. To realize instance reweighting across scales, we develop two complementary regimes: Instance Table, which learns explicit per-sample weights and excels on small/medium data, and Instance Net, a lightweight neural network that generalizes better and scales to large corpora. A practical, stable training recipe -- time-scale matching between upper/lower updates, cold-start initialization, and bounded-range weights -- prevents divergence. Integrated with test-time scaling, DreamPRM-1.5 attains 84.6 accuracy on the MMMU validation set, 31.3 accuracy on R-Bench-V and, when paired with a leading backbone (e.g., GPT-5-mini), achieves first-place results on public multimodal reasoning leaderboards. Moreover, extensive experiments, including benchmark evaluations, baseline comparisons, and a sanity check, demonstrate that DreamPRM-1.5 closes the gap toward the oracle, achieves leading performance, and trains stably. △ Less

Submitted 21 October, 2025; v1 submitted 5 September, 2025; originally announced September 2025.

ERank: Fusing Supervised Fine-Tuning and Reinforcement Learning for Effective and Efficient Text Reranking

Authors: Yuzheng Cai, Yanzhao Zhang, Dingkun Long, Mingxin Li, Pengjun Xie, Weiguo Zheng

Abstract: Text reranking models are a crucial component in modern systems like Retrieval-Augmented Generation, tasked with selecting the most relevant documents prior to generation. However, current Large Language Models (LLMs) powered rerankers often face a fundamental trade-off. On one hand, Supervised Fine-Tuning based pointwise methods that frame relevance as a binary classification task lack the necess… ▽ More Text reranking models are a crucial component in modern systems like Retrieval-Augmented Generation, tasked with selecting the most relevant documents prior to generation. However, current Large Language Models (LLMs) powered rerankers often face a fundamental trade-off. On one hand, Supervised Fine-Tuning based pointwise methods that frame relevance as a binary classification task lack the necessary scoring discrimination, particularly for those built on reasoning LLMs. On the other hand, approaches designed for complex reasoning often employ powerful yet inefficient listwise formulations, rendering them impractical for low latency applications. To resolve this dilemma, we introduce ERank, a highly effective and efficient pointwise reranker built from a reasoning LLM that excels across diverse relevance scenarios. We propose a novel two-stage training pipeline that begins with Supervised Fine-Tuning (SFT). In this stage, we move beyond binary labels and train the model generatively to output fine grained integer scores, which significantly enhances relevance discrimination. The model is then further refined using Reinforcement Learning (RL) with a novel, listwise derived reward. This technique instills global ranking awareness into the efficient pointwise architecture. We evaluate the ERank reranker on the BRIGHT, FollowIR, TREC DL, and BEIR benchmarks, demonstrating superior effectiveness and robustness compared to existing approaches. On the reasoning-intensive BRIGHT benchmark, our ERank-4B achieves an nDCG@10 of 38.7, while a larger 32B variant reaches a state of the art nDCG@10 of 40.2. △ Less

Submitted 30 August, 2025; originally announced September 2025.

Testing and Fault Tolerance Techniques for CNT-Based FPGAs

Authors: Siyuan Lu, Kangwei Xu, Peng Xie, Rui Wang, Yuanqing Cheng

Abstract: As the semiconductor manufacturing process technology node shrinks into the nanometer-scale, the CMOS-based Field Programmable Gate Arrays (FPGAs) face big challenges in scalability of performance and power consumption. Multi-walled Carbon Nanotube (MWCNT) serves as a promising candidate for Cu interconnects thanks to the superior conductivity. Moreover, Carbon Nanotube Field Transistor (CNFET) al… ▽ More As the semiconductor manufacturing process technology node shrinks into the nanometer-scale, the CMOS-based Field Programmable Gate Arrays (FPGAs) face big challenges in scalability of performance and power consumption. Multi-walled Carbon Nanotube (MWCNT) serves as a promising candidate for Cu interconnects thanks to the superior conductivity. Moreover, Carbon Nanotube Field Transistor (CNFET) also emerges as a prospective alternative to the conventional CMOS device because of high power efficiency and large noise margin. The combination of MWCNT and CNFET enables the promising CNT-based FPGAs. However, the MWCNT interconnects exhibit significant process variations due to immature fabrication process, leading to delay faults. Also, the non-ideal CNFET fabrication process may generate a few metallic CNTs (m-CNTs), rendering correlated faulty blocks. In this article, we propose a ring oscillator (RO) based testing technique to detect delay faults due to the process variation of MWCNT interconnects. Furthermore, we propose an effective testing technique for the carry chains in CLBs, and an improved circuit design based on the lookup table (LUT) is applied to speed up the fault testing of CNT-based FPGAs. In addition, we propose a testing algorithm to detect m-CNTs in CLBs. Finally, we propose a redundant spare row sharing architecture to improve the yield of CNT-based FPGA further. Experimental results show that the test time for a 6-input LUT can be reduced by 35.49% compared with conventional testing, and the proposed algorithm can achieve a high test coverage with little overhead. The proposed redundant architecture can repair the faulty segment effectively and efficiently. △ Less

Submitted 18 September, 2025; v1 submitted 27 August, 2025; originally announced August 2025.

Comments: 13 pages

Objective Value Change and Shape-Based Accelerated Optimization for the Neural Network Approximation

Authors: Pengcheng Xie, Zihao Zhou, Zijian Zhou

Abstract: This paper introduce a novel metric of an objective function f, we say VC (value change) to measure the difficulty and approximation affection when conducting an neural network approximation task, and it numerically supports characterizing the local performance and behavior of neural network approximation. Neural networks often suffer from unpredictable local performance, which can hinder their re… ▽ More This paper introduce a novel metric of an objective function f, we say VC (value change) to measure the difficulty and approximation affection when conducting an neural network approximation task, and it numerically supports characterizing the local performance and behavior of neural network approximation. Neural networks often suffer from unpredictable local performance, which can hinder their reliability in critical applications. VC addresses this issue by providing a quantifiable measure of local value changes in network behavior, offering insights into the stability and performance for achieving the neural-network approximation. We investigate some fundamental theoretical properties of VC and identified two intriguing phenomena in neural network approximation: the VC-tendency and the minority-tendency. These trends respectively characterize how pointwise errors evolve in relation to the distribution of VC during the approximation process.In addition, we propose a novel metric based on VC, which measures the distance between two functions from the perspective of variation. Building upon this metric, we further propose a new preprocessing framework for neural network approximation. Numerical results including the real-world experiment and the PDE-related scientific problem support our discovery and pre-processing acceleration method. △ Less

Submitted 27 August, 2025; originally announced August 2025.

Comments: 27 pages

MSC Class: 68T07; 65K05; 65D15; 90C30

InfinityHuman: Towards Long-Term Audio-Driven Human

Authors: Xiaodi Li, Pan Xie, Yi Ren, Qijun Gan, Chen Zhang, Fangyuan Kong, Xiang Yin, Bingyue Peng, Zehuan Yuan

Abstract: Audio-driven human animation has attracted wide attention thanks to its practical applications. However, critical challenges remain in generating high-resolution, long-duration videos with consistent appearance and natural hand motions. Existing methods extend videos using overlapping motion frames but suffer from error accumulation, leading to identity drift, color shifts, and scene instability.… ▽ More Audio-driven human animation has attracted wide attention thanks to its practical applications. However, critical challenges remain in generating high-resolution, long-duration videos with consistent appearance and natural hand motions. Existing methods extend videos using overlapping motion frames but suffer from error accumulation, leading to identity drift, color shifts, and scene instability. Additionally, hand movements are poorly modeled, resulting in noticeable distortions and misalignment with the audio. In this work, we propose InfinityHuman, a coarse-to-fine framework that first generates audio-synchronized representations, then progressively refines them into high-resolution, long-duration videos using a pose-guided refiner. Since pose sequences are decoupled from appearance and resist temporal degradation, our pose-guided refiner employs stable poses and the initial frame as a visual anchor to reduce drift and improve lip synchronization. Moreover, to enhance semantic accuracy and gesture realism, we introduce a hand-specific reward mechanism trained with high-quality hand motion data. Experiments on the EMTD and HDTF datasets show that InfinityHuman achieves state-of-the-art performance in video quality, identity preservation, hand accuracy, and lip-sync. Ablation studies further confirm the effectiveness of each module. Code will be made public. △ Less

Submitted 27 August, 2025; originally announced August 2025.

Comments: Project Page: https://infinityhuman.github.io/

Memp: Exploring Agent Procedural Memory

Authors: Runnan Fang, Yuan Liang, Xiaobin Wang, Jialong Wu, Shuofei Qiao, Pengjun Xie, Fei Huang, Huajun Chen, Ningyu Zhang

Abstract: Large Language Models (LLMs) based agents excel at diverse tasks, yet they suffer from brittle procedural memory that is manually engineered or entangled in static parameters. In this work, we investigate strategies to endow agents with a learnable, updatable, and lifelong procedural memory. We propose Memp that distills past agent trajectories into both fine-grained, step-by-step instructions and… ▽ More Large Language Models (LLMs) based agents excel at diverse tasks, yet they suffer from brittle procedural memory that is manually engineered or entangled in static parameters. In this work, we investigate strategies to endow agents with a learnable, updatable, and lifelong procedural memory. We propose Memp that distills past agent trajectories into both fine-grained, step-by-step instructions and higher-level, script-like abstractions, and explore the impact of different strategies for Build, Retrieval, and Update of procedural memory. Coupled with a dynamic regimen that continuously updates, corrects, and deprecates its contents, this repository evolves in lockstep with new experience. Empirical evaluation on TravelPlanner and ALFWorld shows that as the memory repository is refined, agents achieve steadily higher success rates and greater efficiency on analogous tasks. Moreover, procedural memory built from a stronger model retains its value: migrating the procedural memory to a weaker model yields substantial performance gains. △ Less

Submitted 13 August, 2025; v1 submitted 8 August, 2025; originally announced August 2025.

Comments: Work in progress

WebWatcher: Breaking New Frontier of Vision-Language Deep Research Agent

Authors: Xinyu Geng, Peng Xia, Zhen Zhang, Xinyu Wang, Qiuchen Wang, Ruixue Ding, Chenxi Wang, Jialong Wu, Yida Zhao, Kuan Li, Yong Jiang, Pengjun Xie, Fei Huang, Jingren Zhou

Abstract: Web agents such as Deep Research have demonstrated superhuman cognitive abilities, capable of solving highly challenging information-seeking problems. However, most research remains primarily text-centric, overlooking visual information in the real world. This makes multimodal Deep Research highly challenging, as such agents require much stronger reasoning abilities in perception, logic, knowledge… ▽ More Web agents such as Deep Research have demonstrated superhuman cognitive abilities, capable of solving highly challenging information-seeking problems. However, most research remains primarily text-centric, overlooking visual information in the real world. This makes multimodal Deep Research highly challenging, as such agents require much stronger reasoning abilities in perception, logic, knowledge, and the use of more sophisticated tools compared to text-based agents. To address this limitation, we introduce WebWatcher, a multi-modal Agent for Deep Research equipped with enhanced visual-language reasoning capabilities. It leverages high-quality synthetic multimodal trajectories for efficient cold start training, utilizes various tools for deep reasoning, and further enhances generalization through reinforcement learning. To better evaluate the capabilities of multimodal agents, we propose BrowseComp-VL, a benchmark with BrowseComp-style that requires complex information retrieval involving both visual and textual information. Experimental results show that WebWatcher significantly outperforms proprietary baseline, RAG workflow and open-source agents in four challenging VQA benchmarks, which paves the way for solving complex multimodal information-seeking tasks. △ Less

Submitted 31 August, 2025; v1 submitted 7 August, 2025; originally announced August 2025.

NVSpeech: An Integrated and Scalable Pipeline for Human-Like Speech Modeling with Paralinguistic Vocalizations

Authors: Huan Liao, Qinke Ni, Yuancheng Wang, Yiheng Lu, Haoyue Zhan, Pengyuan Xie, Qiang Zhang, Zhizheng Wu

Abstract: Paralinguistic vocalizations-including non-verbal sounds like laughter and breathing, as well as lexicalized interjections such as "uhm" and "oh"-are integral to natural spoken communication. Despite their importance in conveying affect, intent, and interactional cues, such cues remain largely overlooked in conventional automatic speech recognition (ASR) and text-to-speech (TTS) systems. We presen… ▽ More Paralinguistic vocalizations-including non-verbal sounds like laughter and breathing, as well as lexicalized interjections such as "uhm" and "oh"-are integral to natural spoken communication. Despite their importance in conveying affect, intent, and interactional cues, such cues remain largely overlooked in conventional automatic speech recognition (ASR) and text-to-speech (TTS) systems. We present NVSpeech, an integrated and scalable pipeline that bridges the recognition and synthesis of paralinguistic vocalizations, encompassing dataset construction, ASR modeling, and controllable TTS. (1) We introduce a manually annotated dataset of 48,430 human-spoken utterances with 18 word-level paralinguistic categories. (2) We develop the paralinguistic-aware ASR model, which treats paralinguistic cues as inline decodable tokens (e.g., "You're so funny [Laughter]"), enabling joint lexical and non-verbal transcription. This model is then used to automatically annotate a large corpus, the first large-scale Chinese dataset of 174,179 utterances (573 hours) with word-level alignment and paralingustic cues. (3) We finetune zero-shot TTS models on both human- and auto-labeled data to enable explicit control over paralinguistic vocalizations, allowing context-aware insertion at arbitrary token positions for human-like speech synthesis. By unifying the recognition and generation of paralinguistic vocalizations, NVSpeech offers the first open, large-scale, word-level annotated pipeline for expressive speech modeling in Mandarin, integrating recognition and synthesis in a scalable and controllable manner. Dataset and audio demos are available at https://nvspeech170k.github.io/. △ Less

Submitted 6 August, 2025; originally announced August 2025.

Amber Pruner: Leveraging N:M Activation Sparsity for Efficient Prefill in Large Language Models

Authors: Tai An, Ruwu Cai, Yanzhe Zhang, Yang Liu, Hao Chen, Pengcheng Xie, Sheng Chang, Yiwu Yao, Gongyi Wang

Abstract: In the era of large language models (LLMs), N:M sparsity has emerged as a structured compression technique critical for accelerating inference. While prior work has primarily focused on weight sparsity, it often suffers from significant accuracy degradation. Activation sparsity, though promising, is typically training-dependent and faces challenges in generalization. To address these limitations,… ▽ More In the era of large language models (LLMs), N:M sparsity has emerged as a structured compression technique critical for accelerating inference. While prior work has primarily focused on weight sparsity, it often suffers from significant accuracy degradation. Activation sparsity, though promising, is typically training-dependent and faces challenges in generalization. To address these limitations, we introduce Amber Pruner, a training-free N:M activation sparsity method designed specifically for the prefill stage, targeting the acceleration of linear projection layers in LLMs. Extensive experiments across multiple models and sparsity ratios (2:4, 4:8, and 8:16) demonstrate that Amber Pruner can effectively sparsify and accelerate more than 55% of linear computations without requiring model retraining. To further enhance generality and efficiency, we propose Outstanding-sparse, a unified framework that integrates Amber Pruner with post-training W8A8 quantization. Our approach preserves strong performance across a range of downstream tasks, with notable advantages in generative tasks. This work pioneers a new frontier in activation sparsity, providing foundational insights that are poised to guide the co-evolution of algorithms and architectures in the design of next-generation AI systems. △ Less

Submitted 4 August, 2025; originally announced August 2025.