Showing 1–50 of 336 results for author: Ren, M

Formal Safety Verification for Nonlinear Systems with Generative Barrier Certificate

Authors: Mengxin Ren, Hanrui Zhao

Abstract: Safety verification is a fundamental problem in control theory. Barrier certificates (BCs) provide a powerful formal mechanism, yet deriving BCs is computationally intensive. This paper introduces a generative framework that leverages large language models (LLMs) to synthesize BCs through reasoning. Based on the classical Sum-of-Squares (SOS) approach, we train a domain-specific LLM capable of gen… ▽ More Safety verification is a fundamental problem in control theory. Barrier certificates (BCs) provide a powerful formal mechanism, yet deriving BCs is computationally intensive. This paper introduces a generative framework that leverages large language models (LLMs) to synthesize BCs through reasoning. Based on the classical Sum-of-Squares (SOS) approach, we train a domain-specific LLM capable of generating high-quality BC candidates for nonlinear systems. Then, the LLM-generated BCs transform the intractable Bilinear Matrix Inequality (BMI) solving problems into convex Linear Matrix Inequality (LMI) feasibility test, significantly improving efficiency while preserving correctness. Experimental results show that our generative method achieves several orders of magnitude speedup over traditional numerical BC approaches and, perhaps surprisingly, surpasses the state-of-the-art dedicated neural BC model. These findings mark a substantive step toward integrating generative AI with formal safety verification for dynamical systems. △ Less

Submitted 13 July, 2026; originally announced August 2026.

The finite basis problem for the power semirings of finite groups

Authors: Zidong Gao, Miaomiao Ren, Xiaolei Shao, Mengya Yue

Abstract: For any group $G$, the set of all nonempty subsets of $G$ forms an additively idempotent semiring under set-theoretic union and elementwise multiplication, called the power semiring of $G$ and denoted by $\mathcal{P}(G)$. We prove that for a finite group $G$, $\mathcal{P}(G)$ has no finite basis for its identities if and only if $|G| \geq 3$. This completes the classification of the power semiring… ▽ More For any group $G$, the set of all nonempty subsets of $G$ forms an additively idempotent semiring under set-theoretic union and elementwise multiplication, called the power semiring of $G$ and denoted by $\mathcal{P}(G)$. We prove that for a finite group $G$, $\mathcal{P}(G)$ has no finite basis for its identities if and only if $|G| \geq 3$. This completes the classification of the power semirings of finite groups with respect to the finite basis property. △ Less

Submitted 16 August, 2026; originally announced August 2026.

Comments: 9 pages

The finite basis problem for the flat semirings $S(W)$

Authors: Zidong Gao, Miaomiao Ren, Xianzhong Zhao

Abstract: We focus on the finite basis problem for flat semirings of the form $S(W)$, where $W$ is an arbitrary set of nonempty words. We prove that $S(W)$ generates a Cross variety (and hence is finitely based) whenever every word in $W$ has length at most $3$, whereas it is nonfinitely based whenever there exists $k \geq 3$ such that $W$ is $x^{k+2}$-free but not $x^{k+1}$-free. In particular, if $W_k$ de… ▽ More We focus on the finite basis problem for flat semirings of the form $S(W)$, where $W$ is an arbitrary set of nonempty words. We prove that $S(W)$ generates a Cross variety (and hence is finitely based) whenever every word in $W$ has length at most $3$, whereas it is nonfinitely based whenever there exists $k \geq 3$ such that $W$ is $x^{k+2}$-free but not $x^{k+1}$-free. In particular, if $W_k$ denotes the set of all words of length $k$, then $S(W_k)$ is finitely based if and only if $k \leq 3$. Moreover, $S(W)$ is nonfinitely based whenever $W$ is finite and not $x^4$-free. These results provide a partial answer to an open problem raised by Jackson et al.~(J Algebra 611: 211--245, 2022). △ Less

Submitted 12 August, 2026; originally announced August 2026.

Comments: 26 pages

Free energy of Ising models under a spectral condition

Authors: Andrea Montanari, Michael Ren

Abstract: A sequence of sparse weighted graphs $(G_N:N\ge 1)$ indexed by the number of vertices $N$ is said to be left-convergent if all (suitably weighted) subgraph counts converge to a limit as $N\to\infty$. This notion generalizes in a natural way Benjamini-Schramm's definition of local weak convergence. A broad research agenda aims at determining which `global' graph properties are determined by left or… ▽ More A sequence of sparse weighted graphs $(G_N:N\ge 1)$ indexed by the number of vertices $N$ is said to be left-convergent if all (suitably weighted) subgraph counts converge to a limit as $N\to\infty$. This notion generalizes in a natural way Benjamini-Schramm's definition of local weak convergence. A broad research agenda aims at determining which `global' graph properties are determined by left or local weak convergence (in other words, which of these properties are in fact local). We prove that, under a spectral condition on the weighted adjacency matrix ${\boldsymbol A}_N$ of graph $G_N$, the free energy density of the Ising model on this weighted graph is continuous in the left convergence topology (and hence is a local function). Our proof uses the decomposition of the Ising measure as a log-concave combination of product measures, and of the rapid mixing of Langevin dynamics for log-concave measures. As applications, we derive new limit theorems for the free energy density of spin glasses, antiferromagnets, and magnetization constrained ferromagnetic models, on locally tree-like graphs. △ Less

Submitted 6 August, 2026; originally announced August 2026.

Comments: 41 pages

Can Large Language Models Recover Semantic Optimization Opportunities That Compilers Miss?

Authors: Hailong Jiang, Feng Yu, Emran Hossain, Jianfeng Zhu, Mengfei Ren, Qiang Guan, Chunwei Xia

Abstract: Optimizing compilers miss profitable transformations when their enabling semantics are absent from the analyzed program representation. We ask whether large language models (LLMs) can recover such semantics from heterogeneous C/C++ context and realize them as validated, contract-preserving artifacts. We introduce SeGaBench, an executable benchmark containing 100 synthetic and 20 source-backed case… ▽ More Optimizing compilers miss profitable transformations when their enabling semantics are absent from the analyzed program representation. We ask whether large language models (LLMs) can recover such semantics from heterogeneous C/C++ context and realize them as validated, contract-preserving artifacts. We introduce SeGaBench, an executable benchmark containing 100 synthetic and 20 source-backed cases spanning low-level assumptions, data-structure invariants, and high-level semantic lifting. Each case includes hidden enabling semantics, an oracle artifact, correctness and semantic validators, and a reproducible performance protocol. We evaluate five LLMs using five independent responses per case. The strongest model produces correct artifacts in 94.8% of responses, achieves at least 1.05x speedup in 83.3%, and obtains a performance success on 93.3% of cases. Nevertheless, correct artifacts often close only part of the oracle gap. These results show that LLMs can complement compiler analysis as speculative semantic proposers, provided that their artifacts are validated and evaluated. △ Less

Submitted 4 August, 2026; originally announced August 2026.

Comments: 9 pages, 3 figures

Double shuffle relations imply the infinitesimal hexagon equation

Authors: Muze Ren

Abstract: Double shuffle Lie algebra $\mathfrak{dmr}_0$ was introduced by G.~Racinet in the algebraic study of the multiple zeta values. In this note, we prove that for any $ψ\in \mathfrak{dmr}_0$, it satisfies the infinitesimal hexagon equation $[ψ(x,y),x]+[ψ(-x-y,y),-x-y]=0$. The proof is through the comparison of two different Hopf algebras. Double shuffle Lie algebra $\mathfrak{dmr}_0$ was introduced by G.~Racinet in the algebraic study of the multiple zeta values. In this note, we prove that for any $ψ\in \mathfrak{dmr}_0$, it satisfies the infinitesimal hexagon equation $[ψ(x,y),x]+[ψ(-x-y,y),-x-y]=0$. The proof is through the comparison of two different Hopf algebras. △ Less

Submitted 30 July, 2026; originally announced July 2026.

Comments: 7 pages, comments are welcome

BMOA: Baseline-Mechanism-Outcome Attribution for Compiler-Induced Numerical Deviations

Authors: Hailong Jiang, Emran Hossain, Feng Yu, Chunwei Xia, Mengfei Ren, Jianfeng Zhu, Qiang Guan

Abstract: Formalizing compiler-aware numerical correctness requires distinguishing what an observed floating-point difference means, what compiler behavior the evidence supports, and what numerical consequence follows. Existing testing workflows often collapse these questions into a pass/fail mismatch. We introduce Baseline--Mechanism--Outcome Attribution (BMOA), a diagnostic framework that separates the co… ▽ More Formalizing compiler-aware numerical correctness requires distinguishing what an observed floating-point difference means, what compiler behavior the evidence supports, and what numerical consequence follows. Existing testing workflows often collapse these questions into a pass/fail mismatch. We introduce Baseline--Mechanism--Outcome Attribution (BMOA), a diagnostic framework that separates the comparison relation and system boundary, the evidence-supported compiler mechanism, and the reference-qualified accuracy outcome. BMOA combines operational strict floating-point, transformation-local, reproducibility, cross-compiler, and higher-precision comparisons, while preserving mixed, ambiguous, and unknown attributions when evidence is insufficient. Each record retains inputs, configurations, numerical metrics, and supporting artifacts for audit. We evaluate BMOA on six scientific-computing kernels, deterministic stress-input families, and controlled Clang configurations on ARM64. A 1,276-record attribution corpus and a 162-instance controlled mechanism matrix show that baseline choice changes diagnoses, compiler-induced deviation does not imply accuracy loss, and cancellation and large dynamic range expose the strongest effects within the targeted matrix. BMOA converts raw mismatches into explicit, auditable, evidence-bounded records. Although it is not itself a proof system, these records provide an empirical foundation for future formal specifications and proof obligations for compiler-aware numerical correctness. △ Less

Submitted 29 July, 2026; originally announced July 2026.

RESTOR: Automated Test Oracle Generation for RESTful APIs via Reinforcement Learning

Authors: Xun Zhou, Zhen Dong, Mingyu Ren, Qiang Li, JunJie Li, Sifan Wang, Xiaolong Yu, Chaofeng Sha, Xin Peng

Abstract: Modern REST API testing faces a critical challenge in defining reliable test oracles, particularly in agile industrial environments where formal specifications (e.g., OpenAPI) are frequently missing or outdated, and historical execution logs are unavailable for newly deployed endpoints. In this paper, we present Restor (Reinforcement Enhanced Single-Traffic Oracle generator for REST APIs), a frame… ▽ More Modern REST API testing faces a critical challenge in defining reliable test oracles, particularly in agile industrial environments where formal specifications (e.g., OpenAPI) are frequently missing or outdated, and historical execution logs are unavailable for newly deployed endpoints. In this paper, we present Restor (Reinforcement Enhanced Single-Traffic Oracle generator for REST APIs), a framework that generates executable test assertions from a single observed request-response pair in a black-box setting. Unlike existing approaches that rely on rule-based templates or massive training logs, Restor utilizes a novel data augmentation pipeline to fine-tune a lightweight Large Language Model (LLM) via Group Relative Policy Optimization (GRPO). This training process enables the model to internalize testing "common sense" by optimizing a reward function that jointly encourages: (i) the selection of stable, semantically meaningful fields for validation and the avoidance of dynamic noise (e.g., timestamps or trace IDs); (ii) the generation of robust assertions that withstand logic variations. We evaluate Restor on an industrial dataset comprising over 2,300 API traces across 246 real-world services. Comprehensive experiments demonstrate that Restor significantly outperforms prompt-engineered baselines and generalist models, achieving a superior $F_1$ score of 85.42% in key field identification and increasing the proportion of semantically accurate assertions. Furthermore, deployment in a production CI/CD workflow at ByteDance confirms its practical value: the system raised the adoption rate of automatically generated test cases from 74.1% to over 96%, substantially reducing manual Quality Assurance (QA) effort while ensuring high execution stability. △ Less

Submitted 26 July, 2026; originally announced July 2026.

Comments: Accepted to ISSTA 2026. 22 pages, 8 figures, 2 tables

ACM Class: D.2.5; D.2.11

Agent-Guided Relational Concept Discovery: Toward Interpretable Surgical Margin Assessment

Authors: Nooshin Maghsoodi, Amoon Jamzad, Robert Policelli, Mohammad Farahmand, Dilakshan Srikanthan, Martin Kaufmann, Kevin Y. M. Ren, Shaila Merchant, Sonal Varma, Ross Walker, Doug McKay, John Rudan, Gabor Fichtinger, Parvin Mousavi

Abstract: Deep learning models can effectively use Rapid Evaporative Ionization Mass Spectrometry (REIMS) data for surgical margin assessment. However, their clinical adoption remains challenging due to limited generalization to operating room conditions. This difficulty arises because models are typically trained on labeled spectra collected from resected tissue samples, while they must operate on noisy, u… ▽ More Deep learning models can effectively use Rapid Evaporative Ionization Mass Spectrometry (REIMS) data for surgical margin assessment. However, their clinical adoption remains challenging due to limited generalization to operating room conditions. This difficulty arises because models are typically trained on labeled spectra collected from resected tissue samples, while they must operate on noisy, unlabeled data acquired directly during surgery. In addition, the black-box nature of deep learning models makes it difficult to understand and systematically improve their behavior. Concept-based learning offers a promising way to address these challenges by mapping raw measurements to human-understandable concepts. However, supervised concept-based approaches rely on concept annotations, which are difficult to obtain in complex mass spectrometry workflows. We propose Agent-Guided Concept Discovery, a framework that learns meaningful concepts directly from data without requiring predefined concept labels. During training, a reasoning agent refines semantic descriptions of the learned concepts and adaptively adjusts their weight based on diagnostic relevance. These concepts are further grounded using a biochemical knowledge graph to ensure consistency with known metabolic relationships. Across Skin and Breast Cancer datasets, our model improves balanced accuracy and sensitivity over the baseline. In a representative intraoperative case, it shows fewer false positives, indicating better generalization to surgical conditions. △ Less

Submitted 23 July, 2026; originally announced July 2026.

Comments: This paper is accepted to MICCAI 2026, and this is the submission version, not the camera-ready version

SafeRelBench: A Spatial-Relation-Aware Benchmark for Process-Level Safety in VLM-Driven Embodied Agents

Authors: Huaigang Yang, Ya Li, Min Ren, Bo Dai, Zhenliang Zhang, Zhaofeng He

Abstract: Vision-language models (VLMs) are increasingly used as the reasoning backbone of embodied agents, enabling robots to interpret visual scenes, follow language instructions, and plan multi-step actions. In household environments, however, safety depends not only on recognizing objects, but also on how actions change the physical scene over time. Existing embodied safety evaluations largely focus on… ▽ More Vision-language models (VLMs) are increasingly used as the reasoning backbone of embodied agents, enabling robots to interpret visual scenes, follow language instructions, and plan multi-step actions. In household environments, however, safety depends not only on recognizing objects, but also on how actions change the physical scene over time. Existing embodied safety evaluations largely focus on static risk recognition, unsafe instruction refusal, or final-state task completion. As a result, process-level safety failures induced by spatial relations such as support, containment, and proximity remain insufficiently studied. To address this gap, we introduce SAFERELBENCH, a spatial-relation-aware safety benchmark with 507 executable evaluation samples, including 248 spatial-relation samples and 259 non-spatial control samples. Using SAFERELBENCH to evaluate seven open- and closed-source VLM-driven embodied agents, we find a substantial gap between task success and process-level safety compliance: models often complete the requested task while violating process-level safety constraints. Unlike prior benchmarks, SAFERELBENCH explicitly tests whether agents satisfy safety conditions before risk-prone actions, making spatial relations a core dimension in embodied safety assessment. More broadly, our results show that safe embodied intelligence requires not only stronger perception and planning, but also reliable reasoning about how object relations shape risk during interaction. △ Less

Submitted 15 July, 2026; originally announced July 2026.

Comments: Preprint. 10 pages, 6 figures, 4 tables

Realization and manipulation of spiral charge density waves in a two-dimensional metal

Authors: Lili Zhou, Ruizi Zhang, Chen Si, Zhaoteng Dong, Mengya Ren, Keru Guo, Can Zhang, Jizheng Wu, Fudi Zhou, Huixia Yang, Yaxin Zhao, Guoyuan Yang, Xiaolong Xu, Yuanxiao Ma, Xiao Kong, Yu Zhang, Yeliang Wang

Abstract: Nearly degenerate charge-density-wave (CDW) states play a central role in the competition among collective phenomena. In real materials, however, these states are often intertwined by disorder, hindering their disentanglement and control. Here we show that strain can lift this near-degeneracy and spatially separate distinct CDW states in NbSe2. Using van der Waals (vdW) interactions, we stabilize… ▽ More Nearly degenerate charge-density-wave (CDW) states play a central role in the competition among collective phenomena. In real materials, however, these states are often intertwined by disorder, hindering their disentanglement and control. Here we show that strain can lift this near-degeneracy and spatially separate distinct CDW states in NbSe2. Using van der Waals (vdW) interactions, we stabilize a micron-scale strain network that produces spatially inhomogeneous strain fields. Within this landscape, the intrinsic 3 * 3 CDW superlattice of pristine NbSe2 transforms into an isolated unidirectional 4 * 1 order under 1D-confined compression, and into a 2 * 2 order under biaxial tension. The 4 * 1 CDW has a multiband origin and exhibits markedly enhanced thermal stability, persisting up to 70 K. At strain-network nodes, it further develops into chiral spiral textures, which can be melted by voltage pulses. These results establish strain as a powerful approach to disentangle, stabilize and manipulate competing electronic orders. △ Less

Submitted 15 July, 2026; originally announced July 2026.

Superconducting proximity effect in a strongly correlated charge-transfer insulator

Authors: Mengya Ren, Yaoyao Chen, Fudi Zhou, Can Zhang, Zhaoteng Dong, Lili Zhou, Quanzhen Zhang, Huixia Yang, Xiaolong Xu, Yuanxiao Ma, Yu Zhang, Yeliang Wang

Abstract: Proximity-induced superconductivity in strongly correlated insulators provides a versatile route for engineering quantum states of matter and artificial systems with tailored functionalities. However, microscopic interplay between superconductivity and correlated insulating states remains poorly understood. Here we use ultralow-temperature scanning tunnelling microscopy (STM) to systemically inves… ▽ More Proximity-induced superconductivity in strongly correlated insulators provides a versatile route for engineering quantum states of matter and artificial systems with tailored functionalities. However, microscopic interplay between superconductivity and correlated insulating states remains poorly understood. Here we use ultralow-temperature scanning tunnelling microscopy (STM) to systemically investigate superconducting proximity effects in a charge-transfer insulator. Via STM tip manipulation, atomically sharp lateral junctions composed of superconducting monolayer H-NbSe2 and charge-transfer insulating monolayer T-NbSe2 are constructed, enabling direct access to tunable coupling regimes. In the weak-coupling regime, there is a robust proximity-induced superconducting gap in T-NbSe2, with a reduced gap value relative to that of H-NbSe2. Upon entering the strong-coupling regime, T-NbSe2 exhibits a superconducting gap comparable to that of H-NbSe2, accompanied by pronounced particle-hole-symmetric in-gap bound states, consistent with Yu-Shiba-Rusinov-like excitations. These findings establish monolayer H/T-NbSe2 lateral junctions as a model platform for elucidating superconducting proximity effects in strongly correlated charge-transfer insulators. △ Less

Submitted 15 July, 2026; originally announced July 2026.

The finite basis problem for matrix semirings $\mathbf{M}_n(S_7)$

Authors: Jun Jiao, Miaomiao Ren

Abstract: We first prove that two matrix semirings $\mathbf{M}_n(S_1)$ and $\mathbf{M}_n(S_2)$ are equationally equivalent whenever additively idempotent semirings $S_1$ and $S_2$ are equationally equivalent. We then prove an embedding theorem for matrix semirings $\mathbf{M}_n(S)$ over an additively idempotent semiring $S$: for all $n \geq 2$, $\mathbf{M}_n(S)$ embeds into $\mathbf{M}_{n+1}(S)$. This yield… ▽ More We first prove that two matrix semirings $\mathbf{M}_n(S_1)$ and $\mathbf{M}_n(S_2)$ are equationally equivalent whenever additively idempotent semirings $S_1$ and $S_2$ are equationally equivalent. We then prove an embedding theorem for matrix semirings $\mathbf{M}_n(S)$ over an additively idempotent semiring $S$: for all $n \geq 2$, $\mathbf{M}_n(S)$ embeds into $\mathbf{M}_{n+1}(S)$. This yields an ascending chain of varieties $\mathsf{V}(\mathbf{M}_2(S)) \leq \mathsf{V}(\mathbf{M}_3(S)) \leq \cdots$, which is strictly ascending when $S$ is the two-element distributive lattice. Finally, we show that every variety in the interval $[\mathsf{V}(S_c(abc)), \mathsf{V}(\mathbf{M}_n(S_7))]$ is nonfinitely based (i.e., has no finite basis for its identities), where $S_c(abc)$ is an eight-element flat semiring and $S_7$ is the unique nonfinitely based three-element additively idempotent semiring. Consequently, $\mathbf{M}_n(S_7)$ is nonfinitely based, yielding an ascending chain $\mathsf{V}(\mathbf{M}_2(S_7)) \leq \mathsf{V}(\mathbf{M}_3(S_7)) \leq \cdots$; moreover, every variety in $[\mathsf{V}(S_7), \mathsf{V}(\mathbf{M}_n(S_7))]$ is also nonfinitely based, and this interval contains at least countably infinitely many distinct varieties. △ Less

Submitted 16 August, 2026; v1 submitted 11 June, 2026; originally announced July 2026.

Open-Ended Scenario Reasoning for Specialist Model Adaptation

Authors: Youcheng Zong, Runda Jia, Ranmeng Lin, Mingxuan Ren, Dakuo He

Abstract: Process industries have accumulated validated specialist models, yet sensor drift, feedstock variation, and regime switching cause these models to degrade systematically in new scenarios. Collecting new labeled data and retraining is costly, while continuing with the original model incurs persistent bias. Existing adaptation methods require modifying model parameters with sufficient labeled data,… ▽ More Process industries have accumulated validated specialist models, yet sensor drift, feedstock variation, and regime switching cause these models to degrade systematically in new scenarios. Collecting new labeled data and retraining is costly, while continuing with the original model incurs persistent bias. Existing adaptation methods require modifying model parameters with sufficient labeled data, making rapid response on deployed systems difficult. Using LLMs as direct predictors risks hallucinations and uncontrollable outputs. Such predictors also cannot incorporate unstructured scenario knowledge from the field. To address these limitations, this article proposes Reasoning-Driven Open Adaptation for Specialist Models (ROAM), a framework that uses LLM world knowledge and reasoning to adapt frozen specialist models to unseen scenarios without retraining. ROAM confines all corrections to a low-dimensional, semantically interpretable latent space. LLM-generated scenario judgments and online observations are fused under a unified probabilistic framework. A risk-constrained mechanism suppresses corrections under unreliable LLM evidence or abrupt scenario shifts and falls back to the original frozen model when evidence is insufficient. Experiments on a mineral thickening process and the public IndPenSim penicillin fermentation dataset show that ROAM reduces MAE by over 20\% in major shift settings such as hidden shifts with only 839 additional parameters and under 0.02\,ms per-step overhead. These results indicate that LLM reasoning can be turned into a conservative adaptation signal for industrial models already in service. △ Less

Submitted 7 July, 2026; originally announced July 2026.

LLM-Guided Task-Semantic Field Factorization for Industrial Process Forecasting

Authors: Youcheng Zong, Runda Jia, Mingxuan Ren, Dakuo He

Abstract: Process industries rely on time-series forecasting and soft sensing to estimate quality variables that are hard to measure online. Labeled data are scarce, operating regimes change frequently, and retraining models or rebuilding alignment pipelines for each scenario is costly. Such settings often provide variable tables and process documents that record variable names, units, physical meanings, an… ▽ More Process industries rely on time-series forecasting and soft sensing to estimate quality variables that are hard to measure online. Labeled data are scarce, operating regimes change frequently, and retraining models or rebuilding alignment pipelines for each scenario is costly. Such settings often provide variable tables and process documents that record variable names, units, physical meanings, and process roles. However, standard time-series backbones usually treat inputs as anonymous numerical columns. Existing text-enhanced methods also rarely make the semantic-logical relations between input variables and the prediction target available to the model within each numerical window. To address this problem, this article proposes Task-Semantic Field Factorization (TSF), a large language model (LLM)-guided framework. TSF builds a task-semantic field from task protocols and variable documents before training and uses the LLM only for offline semantic construction. Online training and inference are handled by conventional time-series backbones. During training and inference, the current numerical window activates variable semantics, so semantic information participates in each prediction and supports adaptation to different prediction targets and operating shifts. Across multiple complex industrial forecasting and delayed soft-sensing tasks, TSF reduces MAE by 3.6\% on average. Across all dataset--backbone pairs, the macro-average reduction is 2.9\%, with a maximum reduction of 24.9\%. It adds only about 0.7--4.3k parameters, with less than 8\,$μ$s/sample of additional online inference overhead. These results show that TSF turns existing process documents into measurable forecasting gains across backbones and semantic generators while remaining lightweight for deployment. △ Less

Submitted 18 July, 2026; v1 submitted 7 July, 2026; originally announced July 2026.

MIBE: Multi-subject Interaction Benchmark and Evaluator for Personalized Image Generation

Authors: Zhihan Chen, Yuhuan Zhao, Yijie Zhu, Xinyu Yao, Mengcong Ren, Suwen Wang, Qiuyang Yin, Yuchen Sun, Qin Wang, Lu Xin

Abstract: Multi-subject personalized image generation requires the precise rendering of all requested reference identities and their specified interactions based on a guiding prompt. However, state-of-the-art models still struggle with this process, frequently omitting subjects, failing to preserve reference appearances, or misattributing interactions. Furthermore, existing metrics designed primarily for si… ▽ More Multi-subject personalized image generation requires the precise rendering of all requested reference identities and their specified interactions based on a guiding prompt. However, state-of-the-art models still struggle with this process, frequently omitting subjects, failing to preserve reference appearances, or misattributing interactions. Furthermore, existing metrics designed primarily for single-subject fidelity cannot reliably capture these errors, suffering severe degradation in ranking separability and failing to align with human preference as the subject count increases. To address this gap, we introduce Multi-subject Interaction Benchmark and Evaluator (MIBE), a unified framework comprising a Multi-subject Interaction Benchmark (MIB) and a Multi-subject Interaction Evaluator (MIE). MIB systematically covers diverse relation types and scene complexities through a decoupled data regime. This consists of a 60K-pair VLM-labeled Silver Set for scalable metric training and a 4K-pair double-blind Human Evaluation Gold Set covering a diverse range of state-of-the-art generators, with the Silver Set reaching 95.1% cross-VLM preference agreement. To demonstrate the utility of this benchmark, we present MIE, a lightweight, reference-conditioned evaluator trained exclusively on the Silver Set with a dual-head ranking and diagnosis objective. MIE exhibits strong cross-generator generalization on the Gold Set, achieving 0.922 overall pairwise accuracy against human preference, including 0.982 on seen generators and 0.884 on unseen generators. By outperforming a broad spectrum of baseline metrics, including CLIP and DINO variants, MIE demonstrates that diagnostic supervision can preserve ranking separability and human alignment where traditional evaluators collapse. △ Less

Submitted 1 July, 2026; originally announced July 2026.

AdaJEPA: An Adaptive Latent World Model

Authors: Ying Wang, Oumayma Bounou, Yann LeCun, Mengye Ren

Abstract: Latent world models enable planning from high-dimensional observations by predicting future states in a compact latent space. However, these models are typically kept frozen at test time: when their predictions become inaccurate, planning can fail, especially under test-time distribution shift. To address this, we propose AdaJEPA, an adaptive latent world model that performs test-time adaptation w… ▽ More Latent world models enable planning from high-dimensional observations by predicting future states in a compact latent space. However, these models are typically kept frozen at test time: when their predictions become inaccurate, planning can fail, especially under test-time distribution shift. To address this, we propose AdaJEPA, an adaptive latent world model that performs test-time adaptation within the closed loop of model predictive control (MPC). After training, AdaJEPA plans and executes the first action chunk, uses the observed next-state transition as a self-supervised adaptation signal, and replans with the updated model. This closed-loop update continuously recalibrates the world model without additional expert demonstrations. Across a range of goal-reaching tasks, AdaJEPA substantially improves planning success with as few as one gradient step per MPC replanning step. △ Less

Submitted 30 June, 2026; originally announced June 2026.

Training-free Cross-domain Few-shot Segmentation via Robust Semantic Representation and Matching

Authors: Sujun Sun, Mingwu Ren, Haofeng Zhang

Abstract: Cross-domain Few-shot Segmentation (CD-FSS) aims to transfer knowledge learned from source domain to distinct target domains, segmenting unseen target classes with only a few annotated samples. Although existing methods have made significant progress, they still rely on training or fine-tuning processes, which incur high computational costs and risk overfitting. We observe that when powerful and g… ▽ More Cross-domain Few-shot Segmentation (CD-FSS) aims to transfer knowledge learned from source domain to distinct target domains, segmenting unseen target classes with only a few annotated samples. Although existing methods have made significant progress, they still rely on training or fine-tuning processes, which incur high computational costs and risk overfitting. We observe that when powerful and general-purpose vision foundation models are incorporated into these methods, their performance shows only marginal improvement or even degrades due to overfitting. To address this, we eliminate trainable parameters and propose a training-free framework to avoid both training overhead and overfitting. Built upon the self-supervised vision encoder DINOv3, our framework addresses cross-domain challenges through three core modules. First, the Semantic-aware Feature Re-fusion (SAFR) module identifies and re-fuses features that emphasize semantic patterns, generating representations with enhanced semantic discriminability. Additionally, the Adaptive Support Enhancement (ASE) module narrows semantic gaps between support and query through robust query information aggregation. Finally, the Hybrid Prototype Matching (HPM) module integrates matching results from diverse prototypes to adapt to varying semantic complexity across domains. Extensive experiments on four target domain datasets demonstrate that our method achieves state-of-the-art performance in CD-FSS without any training. △ Less

Submitted 23 June, 2026; originally announced June 2026.

Comments: Accepted by ECCV 2026

Hierarchical Spatial and Channel Aggregation for Cross-domain Few-shot Segmentation

Authors: Sujun Sun, Mingwu Ren, Haofeng Zhang

Abstract: Cross-domain Few-shot Segmentation (CD-FSS) aims to learn generalizable segmentation capability from abundant annotated samples in the source domain, enabling accurate segmentation of novel classes in the target domain with only a few annotated samples. Existing CD-FSS methods mainly focus on mitigating feature distribution shifts caused by style gaps while ignoring significant differences in clas… ▽ More Cross-domain Few-shot Segmentation (CD-FSS) aims to learn generalizable segmentation capability from abundant annotated samples in the source domain, enabling accurate segmentation of novel classes in the target domain with only a few annotated samples. Existing CD-FSS methods mainly focus on mitigating feature distribution shifts caused by style gaps while ignoring significant differences in class semantic granularity and discriminative attributes across domains, leading to two key degradations in support-query matching: semantic over-alignment and attribute over-alignment. To this end, we propose the Dual Hierarchical Aggregation Network (DHANet), which comprises three key modules. First, the Hierarchical Spatial Aggregation (HSA) module performs multi-scale region aggregation of pixel features along the spatial dimension, generating hierarchical semantic-enhanced features to alleviate semantic over-alignment. Additionally, the HCA module conducts multi-scale attribute aggregation along the channel dimension, generating hierarchical attribute-enhanced features to mitigate attribute over-alignment. Finally, we propose the Online Probabilistic Semantic Bank (OPSB), which progressively constructs and updates class probability distributions from query predictions during inference, and samples multiple pseudo-prototypes as additional support information to mitigate insufficient support. Extensive experiments on four target-domain datasets demonstrate that our method achieves state-of-the-art performance. △ Less

Submitted 23 June, 2026; originally announced June 2026.

Comments: Accepted by ECCV 2026

Continual Visual and Verbal Learning Through a Child's Egocentric Input

Authors: Xiaoyang Jiang, Yanlai Yang, Kenneth A. Norman, Brenden Lake, Mengye Ren

Abstract: Children learn the meanings of words from a continuous, temporally structured stream of egocentric experience. Recent work shows that neural networks can also learn word-referent mappings from a child's egocentric video recordings, but they cycle through the shuffled data for hundreds of epochs, contrasting with how children actually encounter their environment. We introduce BabyCL, a continual mu… ▽ More Children learn the meanings of words from a continuous, temporally structured stream of egocentric experience. Recent work shows that neural networks can also learn word-referent mappings from a child's egocentric video recordings, but they cycle through the shuffled data for hundreds of epochs, contrasting with how children actually encounter their environment. We introduce BabyCL, a continual multimodal learning framework that processes the SAYCam dataset in a single chronological pass, combining streaming visual representation learning with an image-text contrastive objective. BabyCL combines a multi-stage temporal segmentation of the stream with a dual replay buffer that independently manages visual and multimodal histories, and it is jointly trained with three contrastive losses on a shared backbone. Under a matched optimization budget, BabyCL outperforms streaming learning baselines on the SAYCam Labeled-S 4AFC benchmark, substantially narrowing the gap to an upper bound of offline training. Ablations show that the gains are robust to the length of the online temporal segmentation window and the eviction rule of the replay buffer. Together, these results show that meaningful word-referent mappings can emerge under training conditions much closer to a child's actual experience. △ Less

Submitted 3 June, 2026; originally announced June 2026.

Comments: 15 pages, 4 figures

FACT: A Simple and Efficient Framework for Active Finetuning

Authors: Wenshuai Xu, You Song, Yuzhuo Cui, Minjie Ren, Qingjie Liu, Zhenghui Hu

Abstract: The main goal of active finetuning is to improve a pretrained model's performance on a specific task or domain by finetuning it with carefully selected informative or challenging data. Previous research has predominantly focused on the active aspect (i.e., data selection) while uniformly employing full finetuning for model adaptation, which inevitably distorts pretrained features due to distributi… ▽ More The main goal of active finetuning is to improve a pretrained model's performance on a specific task or domain by finetuning it with carefully selected informative or challenging data. Previous research has predominantly focused on the active aspect (i.e., data selection) while uniformly employing full finetuning for model adaptation, which inevitably distorts pretrained features due to distribution shift. This issue becomes particularly pronounced when the model size is large relative to the finetuning data quantity, leading to heightened overfitting risks. To address this critical gap, we formally outline the FiAF task that emphasizes systematic exploration of finetuning methodologies in active learning. We propose FACT, a three-phase hierarchical finetuning framework featuring both efficiency and simplicity, specifically designed for active finetuning scenarios. Our comprehensive experiments span: (1) Three major dataset categories encompassing classic (CIFAR10, CIFAR100, ImageNet-1k), imbalanced (CIFAR10-LT, CIFAR100-LT), and fine-grained (StanfordCars, FGVCAircraft) image classification datasets, each evaluated under 3-5 distinct sampling ratios; (2) Diverse pretrained architectures including Convolutional Neural Network (ConvNeXt), Vision Transformer (ViT), and Vision LSTM (ViL) networks; (3) A systematic investigation of frozen feature augmentation (FroFA) strategies. (4) A comprehensive and rigorous analysis of efficiency and generalizability. The results demonstrate significant improvements with strong generalization and robustness. Notably, under low sampling ratios, our framework achieves remarkable performance gains of over 20% on the ViT model for CIFAR10, CIFAR100, and ImageNet-1k benchmarks. This systematic approach establishes new state-of-the-art performance while maintaining parameter efficiency, proving particularly effective when labeled data is scarce. △ Less

Submitted 1 June, 2026; originally announced June 2026.

Comments: ACCEPTED for publication as a REGULAR paper in the IEEE Transactions on Image Processing (T-IP)

Aligning LLMs with Human Uncertainty: A Beta-Bernoulli Calibrator for LLM Forecasting

Authors: Hui Dai, Ryan Teehan, Parsa Torabian, Mengye Ren

Abstract: Probabilistic forecasting estimates the likelihood of uncertain future events. To improve LLM forecasting, existing methods typically learn from binary outcomes to output verbalized forecasts. However, while aggregated human forecasts contain rich information in both the crowd probability estimate and the degree of agreement among forecasters, how to utilize these signals remains underexplored. To… ▽ More Probabilistic forecasting estimates the likelihood of uncertain future events. To improve LLM forecasting, existing methods typically learn from binary outcomes to output verbalized forecasts. However, while aggregated human forecasts contain rich information in both the crowd probability estimate and the degree of agreement among forecasters, how to utilize these signals remains underexplored. To address this, we propose the Beta-Bernoulli Calibrator (BBC), which converts an initial point estimate forecast from any model into a distribution over event likelihood, using supervision from both binary outcomes and human forecasts. BBC models event likelihood $p \sim \text{Beta}(α, β)$ and outcome $y \sim \text{Bernoulli}(p)$, with the mean as the calibrated point forecast and the variance as the epistemic uncertainty. Our results show that BBC generally provides better calibrated and more accurate forecasts than both traditional post-hoc calibration methods and models fine-tuned specifically for forecasting, while remaining lightweight and having good generalization. We also show that the epistemic uncertainty captured by BBC is a more reliable predictor of forecasting error than verbalized confidence. △ Less

Submitted 26 May, 2026; originally announced May 2026.

Semantic Validation of Packer Identification Tools: Characterization, Repair, and Downstream Impact

Authors: Fangtian Zhong, Zhuoyun Qian, Mengfei Ren, Yili Jiang, Jiaqi Huang, Yunming Pang, Xiuzhen Cheng

Abstract: Packer identification tools are a critical foundation of malware analysis, directly affecting unpacking, behavioral analysis, malware classification, and threat attribution. However, their semantic correctness is rarely validated. In practice, a tool may return a plausible packer label that is nevertheless semantically wrong, leading to failed unpacking and unreliable downstream analysis. This pap… ▽ More Packer identification tools are a critical foundation of malware analysis, directly affecting unpacking, behavioral analysis, malware classification, and threat attribution. However, their semantic correctness is rarely validated. In practice, a tool may return a plausible packer label that is nevertheless semantically wrong, leading to failed unpacking and unreliable downstream analysis. This paper presents a semantic validation framework for testing and repairing packer identification tools. Our key idea is to use unpackers as executable semantic contracts. If a tool predicts a packer family, the corresponding unpacker should recover analyzable program content. This enables automatic test oracles without requiring manually labeled ground truth. Building on this idea, we develop a systematic pipeline for detecting, localizing, and repairing semantic faults in existing packer identification tools. We then conduct the first large-scale empirical study of semantic bugs in eleven open-source packer identification tools and six proprietary VirusTotal tools. Our results reveal that semantic bugs are widespread and recurring, largely due to incomplete signatures and unstable heuristic logic. After repair, packer identification coverage improves by up to 58.6%, and downstream malware classification performance improves by more than 13.6% on average. These findings show that semantic validation of packer identification tools is essential for building trustworthy malware analysis pipelines. △ Less

Submitted 27 May, 2026; v1 submitted 25 May, 2026; originally announced May 2026.

StepAudio 2.5 Technical Report

Authors: Bin Lin, Bo Zhao, Boyong Wu, Chao Yan, Chen Wu, Cheng Yi, Chengyuan Yao, Daijiao Liu, Fei Tian, Feng Tian, Haiyang Sun, Haoyang Zhang, Jiangjie Zhen, Jinglan Gong, Jun Chen, Li Xie, Peilin Li, Peng Yang, Pengfei Tan, Qingjian Lin, Runze Li, Shenghua Hu, Siyi Zhou, Wenwen Qu, Xiangyu Li , et al. (76 additional authors not shown)

Abstract: Unified audio-language modeling has emerged as a prominent trend in modern speech systems, promising to bring the reasoning capabilities of large language models to auditory tasks. However, existing unified foundations often struggle to match the depth of specialized systems across automatic speech recognition (ASR), text-to-speech synthesis (TTS), and realtime spoken interaction. Bridging this ga… ▽ More Unified audio-language modeling has emerged as a prominent trend in modern speech systems, promising to bring the reasoning capabilities of large language models to auditory tasks. However, existing unified foundations often struggle to match the depth of specialized systems across automatic speech recognition (ASR), text-to-speech synthesis (TTS), and realtime spoken interaction. Bridging this gap remains an open challenge. This report presents StepAudio 2.5, a unified audio-language foundation model that matches or exceeds specialized systems across all three capabilities. Rather than treating these tasks as architecturally distinct, we operate on the premise that once text and audio share a multimodal representational space, task specialization becomes a matter of operational regimes: data construction, optimization targets, and decoding constraints. Guided by this insight, we advance the post-training paradigm from standard supervised learning to task-tailored Reinforcement Learning from Human Feedback (RLHF), using it as the primary mechanism to define complex optimization targets. We leverage this RLHF-centric alignment, alongside specialized decoding, to shape a shared backbone into three distinct operational modes. Concretely, the ASR branch advances transcription efficiency via verifiable multi-token decoding; the TTS branch achieves controllable, expressive synthesis through preference-based RLHF and context-rich supervision; and the Realtime branch realizes low-latency, persona-consistent dialogue via generative reward modeling within an RLHF framework. On standard benchmarks, StepAudio 2.5 achieves state-of-the-art results across ASR, TTS, and Realtime, demonstrating that a singular audio-language foundation can successfully internalize the distinct deployment objectives of speech understanding, generation, and live interaction. △ Less

Submitted 22 May, 2026; originally announced May 2026.

Brunnian braids and the inclusion from double shuffle Lie algebra to Kashiwara-Vergne Lie algebra

Authors: Muze Ren

Abstract: It is proved by L.~Schneps that the double shuffle Lie algebra $\mathfrak{dmr}_0$ injects to the Kashiwara-Vergne Lie algebra $\mathfrak{krv}_2$ in \cite{Schneps2012,Schneps2025}. We show that $\mathfrak{dmr}_0$ with the infinitesimal hexagon equation $[x,\varphi(-x,-y,x)]+[y,\varphi(-x-y,y)]=0$ injects to the symmetric Kashiwara-Vergne Lie algebra $\mathfrak{krv}^{\mathrm{sym}}_2$. The proof is b… ▽ More It is proved by L.~Schneps that the double shuffle Lie algebra $\mathfrak{dmr}_0$ injects to the Kashiwara-Vergne Lie algebra $\mathfrak{krv}_2$ in \cite{Schneps2012,Schneps2025}. We show that $\mathfrak{dmr}_0$ with the infinitesimal hexagon equation $[x,\varphi(-x,-y,x)]+[y,\varphi(-x-y,y)]=0$ injects to the symmetric Kashiwara-Vergne Lie algebra $\mathfrak{krv}^{\mathrm{sym}}_2$. The proof is based on the inclusion of brunnian braids group on different genus 0 surfaces which is different from the method of mould calculus in \cite{Schneps2012,Schneps2025}. We generalize the inclusion in two directions, one using lower central series of brunnian Lie algebras and the other is to establish explicit links between the pentagon equation map, the stuffle coproduct, the divergence map and the necklace cobracket. △ Less

Submitted 20 May, 2026; originally announced May 2026.

Comments: 24 pages, comments are welcome

Generative Recursive Reasoning

Authors: Junyeob Baek, Mingyu Jo, Minsu Kim, Mengye Ren, Yoshua Bengio, Sungjin Ahn

Abstract: How should future neural reasoning systems implement extended computation? Recursive Reasoning Models (RRMs) offer a promising alternative to autoregressive sequence extension by performing iterative latent-state refinement with shared transition functions. Yet existing RRMs are largely deterministic, following a single latent trajectory and converging to a single prediction. We introduce Generati… ▽ More How should future neural reasoning systems implement extended computation? Recursive Reasoning Models (RRMs) offer a promising alternative to autoregressive sequence extension by performing iterative latent-state refinement with shared transition functions. Yet existing RRMs are largely deterministic, following a single latent trajectory and converging to a single prediction. We introduce Generative Recursive reAsoning Models (GRAM), a framework that turns recursive latent reasoning into probabilistic multi-trajectory computation. GRAM models reasoning as a stochastic latent trajectory, enabling multiple hypotheses, alternative solution strategies, and inference-time scaling through both recursive depth and parallel trajectory sampling. This yields a latent-variable generative model supporting conditional reasoning via $p_θ(y \mid x)$ and, with fixed or absent inputs, unconditional generation via $p_θ(x)$. Trained with amortized variational inference, GRAM improves over deterministic recurrent and recursive baselines on structured reasoning and multi-solution constraint satisfaction tasks, while demonstrating an unconditional generation capability. https://ahn-ml.github.io/gram-website △ Less

Submitted 20 May, 2026; v1 submitted 19 May, 2026; originally announced May 2026.

Seeking the Unfamiliar but Memorable: Conceptual Creativity as Meta-Learning

Authors: Mengye Ren

Abstract: What does it mean to create a new concept, rather than retrieve a familiar one? Repeatedly sampling a generative model at the same prompt produces variations with similar styles and typical content. We propose that creativity is the production of stimuli that are unfamiliar to an adaptive observer at first sight, but quickly learnable from a few exposures. We formalize this as a Creator-Appraiser… ▽ More What does it mean to create a new concept, rather than retrieve a familiar one? Repeatedly sampling a generative model at the same prompt produces variations with similar styles and typical content. We propose that creativity is the production of stimuli that are unfamiliar to an adaptive observer at first sight, but quickly learnable from a few exposures. We formalize this as a Creator-Appraiser pair: a Creator generates a candidate, an Appraiser adapts to it for a few inner-loop learning steps, and the Appraiser's improvement becomes the reward the Creator optimizes through. We instantiate the framework with diffusion as the Creator, an autoencoder Appraiser on MNIST, and a CLIP Appraiser with a low-rank adapter for natural images. The diffusion model remains frozen with no additional language conditioning; the meta-learning gradient is enough to produce both stylistic variations and concept compositions that the base model does not generate on its own. △ Less

Submitted 15 May, 2026; originally announced May 2026.

Comments: 25 pages

A nonfinitely based additively idempotent semiring of order four

Authors: Mengya Yue, Miaomiao Ren

Abstract: We first establish a sufficient condition for an additively idempotent semiring to be nonfinitely based. As applications, we exhibit several examples of additively idempotent semirings satisfying this condition, including a $4$-element semiring $S_{(4,124)}$ whose additive reduct has two minimal elements and two coatoms. Consequently, these semirings have no finite basis for their identities. We first establish a sufficient condition for an additively idempotent semiring to be nonfinitely based. As applications, we exhibit several examples of additively idempotent semirings satisfying this condition, including a $4$-element semiring $S_{(4,124)}$ whose additive reduct has two minimal elements and two coatoms. Consequently, these semirings have no finite basis for their identities. △ Less

Submitted 14 May, 2026; originally announced May 2026.

Learning from Failures: Correction-Oriented Policy Optimization with Verifiable Rewards

Authors: Mengjie Ren, Jie Lou, Boxi Cao, Xueru Wen, Hongyu Lin, Xianpei Han, Le Sun, Xing Yu, Yaojie Lu

Abstract: Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective paradigm for improving the reasoning capabilities of large language models. However, RLVR training is often hindered by sparse binary rewards and weak credit assignment, resulting in ambiguous optimization signals and underutilization of the useful information embedded in failed trajectories. To address this challeng… ▽ More Reinforcement Learning with Verifiable Rewards (RLVR) has emerged as an effective paradigm for improving the reasoning capabilities of large language models. However, RLVR training is often hindered by sparse binary rewards and weak credit assignment, resulting in ambiguous optimization signals and underutilization of the useful information embedded in failed trajectories. To address this challenge, we propose Correction-Oriented Policy Optimization (CIPO), a simple and effective extension to RLVR that converts on-policy failed trajectories into correction-oriented supervision, without relying on any external signals. By jointly optimizing correction samples derived from the model's own failed attempts together with the standard RLVR objective, CIPO improves learning effectiveness while explicitly enhancing the model's ability to correct its own errors. Extensive experiments across 11 benchmarks spanning mathematical reasoning and code generation demonstrate that CIPO consistently and significantly outperforms strong baselines in both reasoning and correction performance. Moreover, CIPO yields stronger pass@K gains, indicating that it improves the model's intrinsic reasoning capacity rather than merely redistributing probability mass over existing correct answers. △ Less

Submitted 14 May, 2026; originally announced May 2026.

Comments: Work on progress

TLX: Hardware-Native, Evolvable MIMW GPU Compiler for Large-scale Production Environments

Authors: Yue Guan, Hongtao Yu, Peng Chen, Daohang Shi, Karthik Manivannan, Nicholas J Riasanovsky, Manman Ren, Lei Wang, Shane Nay, Partha Kanuparthy, Zaifeng Pan, Zhengding Hu, Yufei Ding

Abstract: Modern GPUs increasingly rely on specialized hardware units and asynchronous coordination mechanisms, so performance depends on orchestrating data movement, tensor-core computation, and synchronization rather than exposing more thread-level parallelism. This creates a programming-model tension: if too much execution structure is hidden, the compiler must catch up to new hardware mechanisms; if too… ▽ More Modern GPUs increasingly rely on specialized hardware units and asynchronous coordination mechanisms, so performance depends on orchestrating data movement, tensor-core computation, and synchronization rather than exposing more thread-level parallelism. This creates a programming-model tension: if too much execution structure is hidden, the compiler must catch up to new hardware mechanisms; if too much is exposed, the burden of orchestration falls back onto the programmer. We present TLX (Triton Low-level Language Extensions), built around MIMW (Multi-Instruction, Multi-Warp), which expresses orchestration at warp-group granularity while preserving Triton's productive blocked programming model for regular computation. TLX realizes this idea as an embedded extension to Triton, exposing explicit interfaces for multi-warp execution, local-memory orchestration, asynchronous operations, and cluster-aware control. Our evaluation shows that TLX supports substantial customization with limited development effort while remaining competitive with state-of-the-art implementations. TLX-authored kernels have been deployed in large-scale training and inference production systems. Our code is open sourced at https://github.com/facebookexperimental/triton. △ Less

Submitted 14 May, 2026; v1 submitted 11 May, 2026; originally announced May 2026.

ExecuTorch -- A Unified PyTorch Solution to Run AI Models On-Device

Authors: Mergen Nachin, Digant Desai, Sicheng Stephen Jia, Chen Lai, Mengwei Liu, Jacob Szwejbka, Raziel Alvarez, RJ Ascani, Dave Bort, Manuel Candales, Andrew Caples, Yanan Cao, Zhengxu Chen, Soumith Chintala, Gregory Comer, Tanvir Islam, Songhao Jia, Tarun Karuturi, Jack Khuu, Abhinay Kukkadapu, Tugsbayasgalan Manlaibaatar, Andrew Or, Kimish Patel, Siddartha Pothapragada, Lucy Qiu , et al. (14 additional authors not shown)

Abstract: Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model conversion or complete reimplementation outside the PyTorch ecosystem where the model was originally authored. We introduce ExecuTorch, a unified PyTorch-native deployment framework for edge AI. ExecuTorch enables seamless… ▽ More Local execution of AI on edge devices is important for low latency and offline operation. However, deploying models on diverse hardware remains fragmented, often requiring model conversion or complete reimplementation outside the PyTorch ecosystem where the model was originally authored. We introduce ExecuTorch, a unified PyTorch-native deployment framework for edge AI. ExecuTorch enables seamless deployment of machine learning models across heterogeneous compute environments. It scales from embedded microcontrollers to complex system-on-chips (SoCs) with dedicated accelerators, powering devices ranging from wearables and smartphones to large compute clusters. ExecuTorch preserves PyTorch semantics while allowing customization, support for optimizations like quantization, and pluggable execution "backends". These features together enable fast experimentation, allowing researchers to validate deployment behavior entirely within PyTorch, bridging the gap between research and production. △ Less

Submitted 5 May, 2026; originally announced May 2026.

ProtDBench: A Unified Benchmark of Protein Binder Design and Evaluation

Authors: Cong Liu, Milong Ren, Jiaqi Guan, Chengyue Gong, Jinyuan Sun, Xinshi Chen, Wenzhi Xiao

Abstract: Recent advances in de novo protein binder design have enabled increasing experimental validation, yet reported in silico metrics remain difficult to interpret or compare across studies due to non-standardized evaluation protocols. We introduce ProtDBench, a standardized and throughput-aware evaluation framework for protein binder design. ProtDBench defines unified benchmark tasks, evaluation proto… ▽ More Recent advances in de novo protein binder design have enabled increasing experimental validation, yet reported in silico metrics remain difficult to interpret or compare across studies due to non-standardized evaluation protocols. We introduce ProtDBench, a standardized and throughput-aware evaluation framework for protein binder design. ProtDBench defines unified benchmark tasks, evaluation protocols, and success criteria, enabling systematic analysis of how evaluation design influences observed performance. Using a large wet-lab annotated dataset, we analyze commonly used structure prediction models as evaluation verifiers, revealing substantial verifier-dependent bias and limited agreement under identical filtering protocols. We then benchmark representative open-source generative binder design methods across ten diverse protein targets under a fixed evaluation protocol. Beyond per-sequence success rates, ProtDBench incorporates throughput-aware metrics based on a fixed 24-hour budget, as well as cluster-level success criteria to account for structural diversity. Together, these results expose systematic differences induced by filtering rules, success definitions, and throughput-aware evaluation between computational efficiency, success rate, and structural diversity. Overall, ProtDBench provides a fair and reproducible evaluation pipeline that supports systematic and controlled comparison of protein binder design methods under realistic evaluation settings. △ Less

Submitted 22 May, 2026; v1 submitted 5 May, 2026; originally announced May 2026.

A-CODE: Fully Atomic Protein Co-Design with Unified Multimodal Diffusion

Authors: Chaoran Cheng, Jiaqi Guan, Milong Ren, Chengyue Gong, Cong Liu, Xinshi Chen, Ge Liu, Wenzhi Xiao

Abstract: We present A-CODE, a fully atomic unified one-stage protein co-design model that simultaneously refines discrete atom types and continuous atom coordinates. Unlike predominant two-stage methods that cascade structure design with amino acid-level sequence design, our approach is fully atomic within a unified multimodal diffusion framework, in which residue identities are inferred solely from atom-l… ▽ More We present A-CODE, a fully atomic unified one-stage protein co-design model that simultaneously refines discrete atom types and continuous atom coordinates. Unlike predominant two-stage methods that cascade structure design with amino acid-level sequence design, our approach is fully atomic within a unified multimodal diffusion framework, in which residue identities are inferred solely from atom-level predictions. Built upon the powerful all-atom architecture, A-CODE achieves superior designability for unconditional protein generation, outperforming all existing one-stage and two-stage design models. For binder design, A-CODE rivals and even outperforms existing state-of-the-art two-stage design models and, compared with the existing one-stage co-design model, achieves a drastic tenfold improvement in success rate on hard tasks. The inherent flexibility of our atomic formulation enables, for the first time, seamless adaptation to non-canonical amino acid (ncAA) modeling. Our fully atomic framework establishes a new, versatile foundation for all-atom generative modeling that can be naturally extended to complex biomolecular systems. △ Less

Submitted 5 May, 2026; originally announced May 2026.

A new limit variety of additively idempotent semirings

Authors: Simin Lyu, Miaomiao Ren, Mengya Yue

Abstract: We establish a sufficient condition for an additively idempotent semiring to be nonfinitely based. Applying this condition, we prove that the six-element additively idempotent semiring $SR_6$ has no finite basis for its identity. Furthermore, we provide a complete description of the subvariety lattice of the variety $\mathsf{V}(SR_6)$ generated by $SR_6$, showing that it forms a four-element chain… ▽ More We establish a sufficient condition for an additively idempotent semiring to be nonfinitely based. Applying this condition, we prove that the six-element additively idempotent semiring $SR_6$ has no finite basis for its identity. Furthermore, we provide a complete description of the subvariety lattice of the variety $\mathsf{V}(SR_6)$ generated by $SR_6$, showing that it forms a four-element chain. Our results demonstrate that $\mathsf{V}(SR_6)$ is a limit variety: it is itself nonfinitely based, yet all of its proper subvarieties are finitely based. Moreover, $SR_6$ is the smallest known example of an additively idempotent semiring generating a limit variety. △ Less

Submitted 13 March, 2026; originally announced April 2026.

CodePivot: Bootstrapping Multilingual Transpilation in LLMs via Reinforcement Learning without Parallel Corpora

Authors: Shangyu Li, Juyong Jiang, Meibo Ren, Sizhe Zhong, Huiri Tan, Yunhao Gou, Xu Han, Chun Yong Chong, Yun Peng, Jiasi Shen

Abstract: Transpilation, or code translation, aims to convert source code from one programming language (PL) to another. It is beneficial for many downstream applications, from modernizing large legacy codebases to augmenting data for low-resource PLs. Recent large language model (LLM)-based approaches have demonstrated immense potential for code translation. Among these approaches, training-based methods a… ▽ More Transpilation, or code translation, aims to convert source code from one programming language (PL) to another. It is beneficial for many downstream applications, from modernizing large legacy codebases to augmenting data for low-resource PLs. Recent large language model (LLM)-based approaches have demonstrated immense potential for code translation. Among these approaches, training-based methods are particularly important because LLMs currently do not effectively adapt to domain-specific settings that suffer from a lack of knowledge without targeted training. This limitation is evident in transpilation tasks involving low-resource PLs. However, existing training-based approaches rely on a pairwise transpilation paradigm, making it impractical to support a diverse range of PLs. This limitation is particularly prominent for low-resource PLs due to a scarcity of training data. Furthermore, these methods suffer from suboptimal reinforcement learning (RL) reward formulations. To address these limitations, we propose CodePivot, a training framework that leverages Python as an intermediate representation (IR), augmented by a novel RL reward mechanism, Aggressive-Partial-Functional reward, to bootstrap the model's multilingual transpilation ability without requiring parallel corpora. Experiments involving 10 PLs show that the resulting 7B model, trained on Python-to-Others tasks, consistently improves performance across both general and low-resource PL-related transpilation tasks. It outperforms substantially larger mainstream models with hundreds of billions more parameters, such as Deepseek-R1 and Qwen3-235B-A22B-Instruct-2507, on Python-to-Others tasks and Others-to-All tasks, respectively. In addition, it outperforms its counterpart trained directly on Any-to-Any tasks on general transpilation tasks. The code and data are available at https://github.com/lishangyu-hkust/CodePivot. △ Less

Submitted 20 April, 2026; originally announced April 2026.

TokenLight: Precise Lighting Control in Images using Attribute Tokens

Authors: Sumit Chaturvedi, Yannick Hold-Geoffroy, Mengwei Ren, Jingyuan Liu, He Zhang, Yiqun Mei, Julie Dorsey, Zhixin Shu

Abstract: This paper presents a method for image relighting that enables precise and continuous control over multiple illumination attributes in a photograph. We formulate relighting as a conditional image generation task and introduce attribute tokens to encode distinct lighting factors such as intensity, color, ambient illumination, diffuse level, and 3D light positions. The model is trained on a large-sc… ▽ More This paper presents a method for image relighting that enables precise and continuous control over multiple illumination attributes in a photograph. We formulate relighting as a conditional image generation task and introduce attribute tokens to encode distinct lighting factors such as intensity, color, ambient illumination, diffuse level, and 3D light positions. The model is trained on a large-scale synthetic dataset with ground-truth lighting annotations, supplemented by a small set of real captures to enhance realism and generalization. We validate our approach across a variety of relighting tasks, including controlling in-scene lighting fixtures and editing environment illumination using virtual light sources, on synthetic and real images. Our method achieves state-of-the-art quantitative and qualitative performance compared to prior work. Remarkably, without explicit inverse rendering supervision, the model exhibits an inherent understanding of how light interacts with scene geometry, occlusion, and materials, yielding convincing lighting effects even in traditionally challenging scenarios such as placing lights within objects or relighting transparent materials plausibly. Project page: vrroom.github.io/tokenlight/ △ Less

Submitted 17 April, 2026; v1 submitted 16 April, 2026; originally announced April 2026.

Comments: 32 pages, CVPR 2026, Project Page: https://vrroom.github.io/tokenlight/

Memory Transfer Learning: How Memories are Transferred Across Domains in Coding Agents

Authors: Kangsan Kim, Minki Kang, Taeil Kim, Yanlai Yang, Mengye Ren, Sung Ju Hwang

Abstract: Memory-based self-evolution has emerged as a promising paradigm for coding agents. However, existing approaches typically restrict memory utilization to homogeneous task domains, failing to leverage the shared infrastructural foundations, such as runtime environments and programming languages, that exist across diverse real-world coding problems. To address this limitation, we investigate \textbf{… ▽ More Memory-based self-evolution has emerged as a promising paradigm for coding agents. However, existing approaches typically restrict memory utilization to homogeneous task domains, failing to leverage the shared infrastructural foundations, such as runtime environments and programming languages, that exist across diverse real-world coding problems. To address this limitation, we investigate \textbf{Memory Transfer Learning} (MTL) by harnessing a unified memory pool from heterogeneous domains. We evaluate performance across 6 coding benchmarks using four memory representations, ranging from concrete traces to abstract insights. Our experiments demonstrate that cross-domain memory improves average performance by 3.7\%, primarily by transferring meta-knowledge, such as validation routines, rather than task-specific code. Importantly, we find that abstraction dictates transferability; high-level insights generalize well, whereas low-level traces often induce negative transfer due to excessive specificity. Furthermore, we show that transfer effectiveness scales with the size of the memory pool, and memory can be transferred even between different models. Our work establishes empirical design principles for expanding memory utilization beyond single-domain silos. Project page: https://memorytransfer.github.io/ △ Less

Submitted 15 April, 2026; originally announced April 2026.

Comments: Preprint

Temporal Straightening for Latent Planning

Authors: Ying Wang, Oumayma Bounou, Gaoyue Zhou, Randall Balestriero, Tim G. J. Rudner, Yann LeCun, Mengye Ren

Abstract: Learning good representations is essential for latent planning with world models. While pretrained visual encoders produce strong semantic visual features, they are not tailored to planning and contain information irrelevant -- or even detrimental -- to planning. Inspired by the perceptual straightening hypothesis in human visual processing, we introduce temporal straightening to improve represent… ▽ More Learning good representations is essential for latent planning with world models. While pretrained visual encoders produce strong semantic visual features, they are not tailored to planning and contain information irrelevant -- or even detrimental -- to planning. Inspired by the perceptual straightening hypothesis in human visual processing, we introduce temporal straightening to improve representation learning for latent planning. Using a curvature regularizer that encourages locally straightened latent trajectories, we jointly learn an encoder and a predictor of a Joint-Embedding Predictive Architecture (JEPA) world model. We show that reducing curvature this way makes the Euclidean distance in latent space a better proxy for the geodesic distance and improves the conditioning of the planning objective. We demonstrate empirically that temporal straightening makes gradient-based planning more stable and yields significantly higher success rates across a suite of goal-reaching tasks. Our code is available at https://agenticlearning.ai/temporal-straightening. △ Less

Submitted 11 August, 2026; v1 submitted 12 March, 2026; originally announced March 2026.

Comments: ICML2026 Camera Ready

Tackling Length Inflation Without Trade-offs: Group Relative Reward Rescaling for Reinforcement Learning

Authors: Zichao Li, Jie Lou, Fangchen Dong, Zhiyuan Fan, Mengjie Ren, Hongyu Lin, Xianpei Han, Debing Zhang, Le Sun, Yaojie Lu, Xing Yu

Abstract: Reinforcement learning significantly enhances LLM capabilities but suffers from a critical issue: length inflation, where models adopt verbosity or inefficient reasoning to maximize rewards. Prior approaches struggle to address this challenge in a general and lossless manner, primarily because additive penalties introduce a compensatory effect that creates optimization shortcuts, while heuristic g… ▽ More Reinforcement learning significantly enhances LLM capabilities but suffers from a critical issue: length inflation, where models adopt verbosity or inefficient reasoning to maximize rewards. Prior approaches struggle to address this challenge in a general and lossless manner, primarily because additive penalties introduce a compensatory effect that creates optimization shortcuts, while heuristic gating strategies lack generality beyond binary feedback. To bridge this gap, we present Group Relative Reward Rescaling (GR$^3$), which reframes length control as a multiplicative rescaling paradigm, effectively establishing a generalized, continuous, and reward-dependent gating mechanism. To further ensure lossless optimization, we incorporate group-relative regularization and advantage-aware calibration, which dynamically adapt length budgets to instance difficulty and preserve the advantage signal of high-quality trajectories. Empirically, across both RLHF and RLVR settings, GR$^3$~maintains training dynamics and downstream performance comparable to standard GRPO while significantly mitigating length inflation, outperforming state-of-the-art length-regularized baselines. △ Less

Submitted 11 March, 2026; originally announced March 2026.

MA-EgoQA: Question Answering over Egocentric Videos from Multiple Embodied Agents

Authors: Kangsan Kim, Yanlai Yang, Suji Kim, Woongyeong Yeo, Youngwan Lee, Mengye Ren, Sung Ju Hwang

Abstract: As embodied models become powerful, humans will collaborate with multiple embodied AI agents at their workplace or home in the future. To ensure better communication between human users and the multi-agent system, it is crucial to interpret incoming information from agents in parallel and refer to the appropriate context for each query. Existing challenges include effectively compressing and commu… ▽ More As embodied models become powerful, humans will collaborate with multiple embodied AI agents at their workplace or home in the future. To ensure better communication between human users and the multi-agent system, it is crucial to interpret incoming information from agents in parallel and refer to the appropriate context for each query. Existing challenges include effectively compressing and communicating high volumes of individual sensory inputs in the form of video and correctly aggregating multiple egocentric videos to construct system-level memory. In this work, we first formally define a novel problem of understanding multiple long-horizon egocentric videos simultaneously collected from embodied agents. To facilitate research in this direction, we introduce MultiAgent-EgoQA (MA-EgoQA), a benchmark designed to systemically evaluate existing models in our scenario. MA-EgoQA provides 1.7k questions unique to multiple egocentric streams, spanning five categories: social interaction, task coordination, theory-of-mind, temporal reasoning, and environmental interaction. We further propose a simple baseline model for MA-EgoQA named EgoMAS, which leverages shared memory across embodied agents and agent-wise dynamic retrieval. Through comprehensive evaluation across diverse baselines and EgoMAS on MA-EgoQA, we find that current approaches are unable to effectively handle multiple egocentric streams, highlighting the need for future advances in system-level understanding across the agents. The code and benchmark are available at https://ma-egoqa.github.io. △ Less

Submitted 10 March, 2026; v1 submitted 10 March, 2026; originally announced March 2026.

Comments: Under review

Adaptive Active Learning for Online Reliability Prediction of Satellite Electronics

Authors: Shixiang Li, Yubin Tian, Dianpeng Wang, Piao Chen, Mengying Ren

Abstract: Accurate on-orbit reliability prediction for satellite electronics is often hindered by limited data availability, varying operational conditions, and considerable unit-to-unit variability. To overcome these obstacles, this paper proposes a novel integrated online reliability prediction framework. The main contributions are twofold. First, a Wiener process-based degradation model is developed, inc… ▽ More Accurate on-orbit reliability prediction for satellite electronics is often hindered by limited data availability, varying operational conditions, and considerable unit-to-unit variability. To overcome these obstacles, this paper proposes a novel integrated online reliability prediction framework. The main contributions are twofold. First, a Wiener process-based degradation model is developed, incorporating a generalized Arrhenius link function, individual random effects, and spatial correlations among adjacent units. A customized maximum likelihood estimation method is further devised to facilitate efficient and accurate parameter inference. Second, a two-stage active learning sampling scheme is designed to adaptively enhance prediction accuracy. This strategy initially selects representative units based on spatial configuration, and subsequently determines optimal sampling times using a comprehensive criterion that balances unit-specific information, model uncertainty, and degradation dynamics. Numerical experiments and a practical case study from the Tiangong space station demonstrate that the proposed method markedly improves reliability prediction accuracy while significantly reducing data requirements, offering an efficient solution for the prognostic and health management of complex satellite electronic systems. △ Less

Submitted 9 March, 2026; originally announced March 2026.

Impacts of Jet Stream Structure on Cyclone Merging and Persistent Anticyclones: Insights from Dry Idealized Simulations

Authors: Mingfei Ren, Gan Zhang, Kai-Yuan Cheng, Lucas Harris, Talia Tamarin-Brodsky, Joseph Mouallem

Abstract: Midlatitude jet streams exhibit substantial variability in latitude, width, and vertical depth on synoptic to multi-decadal timescales. While the upper-level dynamics of baroclinic waves have been extensively studied, the sensitivity of the extreme-generating, low-level phenomena to these variations remains underexplored. Here, we systematically investigate this sensitivity using dry, adiabatic id… ▽ More Midlatitude jet streams exhibit substantial variability in latitude, width, and vertical depth on synoptic to multi-decadal timescales. While the upper-level dynamics of baroclinic waves have been extensively studied, the sensitivity of the extreme-generating, low-level phenomena to these variations remains underexplored. Here, we systematically investigate this sensitivity using dry, adiabatic idealized experiments with the GFDL FV3 dry dynamical core initialized with analytically specified jets. We identify jet variations that control synoptic-scale features of interest. Results indicate that poleward-shifted jets accelerate initial cyclone intensification and favor anticyclonic Rossby Wave Breaking (RWB). These wave-breaking tendencies are consistent with established baroclinic paradigms, validating the newly configured idealized simulations. Additionally, jet width regulates the likelihood of surface cyclone merging. Poleward-shifted, broader, and higher jets produce more frequent cyclone merging, generating intense wind extremes. Finally, we show that poleward-shifted, broad, deep jets dynamically precondition the flow for persistent stationary anticyclones in the absence of diabatic contributions. Together, these findings illustrate how changes in jet stream structure may modulate midlatitude weather extremes. △ Less

Submitted 8 March, 2026; originally announced March 2026.

How Long Can Unified Multimodal Models Generate Images Reliably? Taming Long-Horizon Interleaved Image Generation via Context Curation

Authors: Haoyu Chen, Qing Liu, Yuqian Zhou, He Zhang, Zhaowen Wang, Mengwei Ren, Jingjing Ren, Xiang Wang, Zhe Lin, Lei Zhu

Abstract: Unified multimodal models hold the promise of generating extensive, interleaved narratives, weaving text and imagery into coherent long-form stories. However, current systems suffer from a critical reliability gap: as sequences grow, generation quality rapidly collapses. In this work, we investigate the mechanism behind this failure and argue that it is distinct from standard long-context challeng… ▽ More Unified multimodal models hold the promise of generating extensive, interleaved narratives, weaving text and imagery into coherent long-form stories. However, current systems suffer from a critical reliability gap: as sequences grow, generation quality rapidly collapses. In this work, we investigate the mechanism behind this failure and argue that it is distinct from standard long-context challenges. We reveal that in generation, accumulated visual history acts as a source of active pollution, a decay governed specifically by the number of image events rather than raw token count. We identify a structural vulnerability where dense visual tokens overwhelm the attention mechanism, creating noise that distorts future synthesis. Guided by these mechanistic insights, we propose UniLongGen, a training-free inference strategy that prioritizes safe conditioning over total recall. Instead of retaining all history, UniLongGen dynamically curates the model's memory, identifying and discarding interfering visual signals based on the model's own internal relevance rankings. Extensive experiments demonstrate that this active forgetting approach is essential for stability: UniLongGen significantly outperforms baselines in long-horizon fidelity and consistency, while simultaneously reducing memory footprint and inference time. △ Less

Submitted 8 March, 2026; originally announced March 2026.

One step further with Monte-Carlo sampler to guide diffusion better

Authors: Minsi Ren, Wenhao Deng, Ruiqi Feng, Tailin Wu

Abstract: Stochastic differential equation (SDE)-based generative models have achieved substantial progress in conditional generation via training-free differentiable loss-guided approaches. However, existing methodologies utilizing posterior sam- pling typically confront a substantial estimation error, which results in inaccu- rate gradients for guidance and leading to inconsistent generation results. To m… ▽ More Stochastic differential equation (SDE)-based generative models have achieved substantial progress in conditional generation via training-free differentiable loss-guided approaches. However, existing methodologies utilizing posterior sam- pling typically confront a substantial estimation error, which results in inaccu- rate gradients for guidance and leading to inconsistent generation results. To mitigate this issue, we propose that performing an additional backward denois- ing step and Monte-Carlo sampling (ABMS) can achieve better guided diffu- sion, which is a plug-and-play adjustment strategy. To verify the effectiveness of our method, we provide theoretical analysis and propose the adoption of a dual-focus evaluation framework, which further serves to highlight the critical problem of cross-condition interference prevalent in existing approaches. We conduct experiments across various task settings and data types, mainly includ- ing conditional online handwritten trajectory generation, image inverse problems (inpainting, super resolution and gaussian deblurring) molecular inverse design and so on. Experimental results demonstrate that our approach can be effec- tively used with higher order samplers and consistently improves the quality of generation samples across all the different scenarios. △ Less

Submitted 4 March, 2026; originally announced March 2026.

Comments: 16 pages, 7 figures, accepted at ICLR2026

AI+HW 2035: Shaping the Next Decade

Authors: Deming Chen, Jason Cong, Azalia Mirhoseini, Christos Kozyrakis, Subhasish Mitra, Jinjun Xiong, Cliff Young, Anima Anandkumar, Michael Littman, Aron Kirschen, Sophia Shao, Serge Leef, Naresh Shanbhag, Dejan Milojicic, Michael Schulte, Gert Cauwenberghs, Jerry M. Chow, Tri Dao, Kailash Gopalakrishnan, Richard Ho, Hoshik Kim, Kunle Olukotun, David Z. Pan, Mark Ren, Dan Roth , et al. (5 additional authors not shown)

Abstract: Artificial intelligence (AI) and hardware (HW) are advancing at unprecedented rates, yet their trajectories have become inseparably intertwined. The global research community lacks a cohesive, long-term vision to strategically coordinate the development of AI and HW. This fragmentation constrains progress toward holistic, sustainable, and adaptive AI systems capable of learning, reasoning, and ope… ▽ More Artificial intelligence (AI) and hardware (HW) are advancing at unprecedented rates, yet their trajectories have become inseparably intertwined. The global research community lacks a cohesive, long-term vision to strategically coordinate the development of AI and HW. This fragmentation constrains progress toward holistic, sustainable, and adaptive AI systems capable of learning, reasoning, and operating efficiently across cloud, edge, and physical environments. The future of AI depends not only on scaling intelligence, but on scaling efficiency, achieving exponential gains in intelligence per joule, rather than unbounded compute consumption. Addressing this grand challenge requires rethinking the entire computing stack. This vision paper lays out a 10-year roadmap for AI+HW co-design and co-development, spanning algorithms, architectures, systems, and sustainability. We articulate key insights that redefine scaling around energy efficiency, system-level integration, and cross-layer optimization. We identify key challenges and opportunities, candidly assess potential obstacles and pitfalls, and propose integrated solutions grounded in algorithmic innovation, hardware advances, and software abstraction. Looking ahead, we define what success means in 10 years: achieving a 1000x improvement in efficiency for AI training and inference; enabling energy-aware, self-optimizing systems that seamlessly span cloud, edge, and physical AI; democratizing access to advanced AI infrastructure; and embedding human-centric principles into the design of intelligent systems. Finally, we outline concrete action items for academia, industry, government, and the broader community, calling for coordinated national initiatives, shared infrastructure, workforce development, cross-agency collaboration, and sustained public-private partnerships to ensure that AI+HW co-design becomes a unifying long-term mission. △ Less

Submitted 5 March, 2026; originally announced March 2026.

Comments: 35 pages, 4 figures

TacMamba: A Tactile History Compression Adapter Bridging Fast Reflexes and Slow VLA Reasoning

Authors: Zhenan Wang, Yanzhe Wang, Meixuan Ren, Peng Li, Yang Liu, Yifei Nie, Limin Long, Yun Ye, Xiaofeng Wang, Zhen Zhu, Huixu Dong

Abstract: In visually ambiguous manipulation such as detecting button click tactile feedback is often the sole source of ground truth. However, fusing tactile data poses a significant challenge due to a spatiotemporal mismatch: tactile perception requires high-frequency processing with long-horizon memory (System 1), whereas visual policies operate at low control frequencies (System 2). Existing architectur… ▽ More In visually ambiguous manipulation such as detecting button click tactile feedback is often the sole source of ground truth. However, fusing tactile data poses a significant challenge due to a spatiotemporal mismatch: tactile perception requires high-frequency processing with long-horizon memory (System 1), whereas visual policies operate at low control frequencies (System 2). Existing architectures struggle to bridge this gap: Transformers are computationally prohibitive for high-frequency loops (>100Hz), while LSTMs suffer from forgetting over extended interaction histories. In this paper, we introduce TacMamba, a hierarchical architecture that aligns high-bandwidth tactile reflexes with low-frequency visual planning. Our approach comprises three core contributions: (1) a custom high-frequency tactile interface designed for flexible integration; (2) a Mamba-based Tactile History Compressor that encodes continuous force history into a compact state with O(1) inference latency (0.45 ms), enabling plug-and-play fusion with VLA models without joint pre-training and (3) a Tactile-Guided Dual-Stage Training strategy that leverages temporal discrimination for self-supervised representation learning and phase-uniform sampling to mitigate data sparsity. Experiments on discrete counting and implicit state switching demonstrate that TacMamba achieves 100% success rates, significantly outperforming the visual-only pi_0.5 baseline, while strictly satisfying hard real-time constraints. △ Less

Submitted 2 March, 2026; originally announced March 2026.

Two nonfinitely based additively idempotent semirings of order four

Authors: Mengya Yue, Miaomiao Ren, Zidong Gao

Abstract: We establish two sufficient conditions for an additively idempotent semiring to be nonfinitely based. As applications, we prove that two specific $4$-element additively idempotent semirings, $S_{(4,545)}$ and $S_{(4,634)}$, whose additive reducts are chains, have no finite basis for their identities. Furthermore, we show that the interval $[\mathsf{V}(S_{(4,545)}),\mathsf{V}(S_{(4,634)})]$ in the… ▽ More We establish two sufficient conditions for an additively idempotent semiring to be nonfinitely based. As applications, we prove that two specific $4$-element additively idempotent semirings, $S_{(4,545)}$ and $S_{(4,634)}$, whose additive reducts are chains, have no finite basis for their identities. Furthermore, we show that the interval $[\mathsf{V}(S_{(4,545)}),\mathsf{V}(S_{(4,634)})]$ in the lattice of semiring varieties contains \(2^{\aleph_0}\) distinct varieties. Consequently, the join of two finitely based additively idempotent semiring varieties is not necessarily finitely based. Moreover, we obtain the smallest example of a finitely based additively idempotent semiring $S$ whose extension $S^0$ (obtained by adjoining a new element) is nonfinitely based. △ Less

Submitted 30 January, 2026; originally announced March 2026.

Step 3.5 Flash: Open Frontier-Level Intelligence with 11B Active Parameters

Authors: Ailin Huang, Ang Li, Aobo Kong, Bin Wang, Binxing Jiao, Bo Dong, Bojun Wang, Boyu Chen, Brian Li, Buyun Ma, Chang Su, Changxin Miao, Changyi Wan, Chao Lou, Chen Hu, Chen Xu, Chenfeng Yu, Chengting Feng, Chengyuan Yao, Chunrui Han, Dan Ma, Dapeng Shi, Daxin Jiang, Dehua Ma, Deshan Sun , et al. (191 additional authors not shown)

Abstract: We introduce Step 3.5 Flash, a sparse Mixture-of-Experts (MoE) model that bridges frontier-level agentic intelligence and computational efficiency. We focus on what matters most when building agents: sharp reasoning and fast, reliable execution. Step 3.5 Flash pairs a 196B-parameter foundation with 11B active parameters for efficient inference. It is optimized with interleaved 3:1 sliding-window/f… ▽ More We introduce Step 3.5 Flash, a sparse Mixture-of-Experts (MoE) model that bridges frontier-level agentic intelligence and computational efficiency. We focus on what matters most when building agents: sharp reasoning and fast, reliable execution. Step 3.5 Flash pairs a 196B-parameter foundation with 11B active parameters for efficient inference. It is optimized with interleaved 3:1 sliding-window/full attention and Multi-Token Prediction (MTP-3) to reduce the latency and cost of multi-round agentic interactions. To reach frontier-level intelligence, we design a scalable reinforcement learning framework that combines verifiable signals with preference feedback, while remaining stable under large-scale off-policy training, enabling consistent self-improvement across mathematics, code, and tool use. Step 3.5 Flash demonstrates strong performance across agent, coding, and math tasks, achieving 85.4% on IMO-AnswerBench, 86.4% on LiveCodeBench-v6 (2024.08-2025.05), 88.2% on tau2-Bench, 69.0% on BrowseComp (with context management), and 51.0% on Terminal-Bench 2.0, comparable to frontier models such as GPT-5.2 xHigh and Gemini 3.0 Pro. By redefining the efficiency frontier, Step 3.5 Flash provides a high-density foundation for deploying sophisticated agents in real-world industrial environments. △ Less

Submitted 23 February, 2026; v1 submitted 11 February, 2026; originally announced February 2026.

Comments: Technical report for Step 3.5 Flash

The finite basis problem for matrix semirings over a two-element additively idempotent semiring

Authors: Jun Jiao, Miaomiao Ren

Abstract: We provide a complete classification of matrix semirings $\mathbf{M}_n(S)$ over two-element additively idempotent semirings $S$ with respect to the finite basis property.Our main theorem shows that for every integer $n \geq 2$,the semiring $\mathbf{M}_n(S)$ is finitely based if and only if $S$ is distinct from a distributive lattice. We provide a complete classification of matrix semirings $\mathbf{M}_n(S)$ over two-element additively idempotent semirings $S$ with respect to the finite basis property.Our main theorem shows that for every integer $n \geq 2$,the semiring $\mathbf{M}_n(S)$ is finitely based if and only if $S$ is distinct from a distributive lattice. △ Less

Submitted 11 January, 2026; originally announced February 2026.

Anomalous Non-Hermitian Topological Anderson Insulator

Authors: Mina Ren, Xi Shi, Haitao Jiang, Feng Liu, Hong Chen, Yong Sun

Abstract: Strong disorder drives conventional Hermitian systems into Anderson insulating states, suppressing all topological phases. Here, we unveil symmetry-protected, anomalous topological phases in the strong disorder limit of a non-Hermitian system, characterized by a scale-invariant merging of zero-energy modes. Using the maximally symmetric Jx lattice as an ideal platform and introducing specifically… ▽ More Strong disorder drives conventional Hermitian systems into Anderson insulating states, suppressing all topological phases. Here, we unveil symmetry-protected, anomalous topological phases in the strong disorder limit of a non-Hermitian system, characterized by a scale-invariant merging of zero-energy modes. Using the maximally symmetric Jx lattice as an ideal platform and introducing specifically engineered (ABBA-type) symmetry-preserving non-Hermitian disorder, we observe a sequence of disorder-induced phase transitions: from a trivial insulator into and through a non-Hermitian topological Anderson insulator (TAI) phase, culminating in a stable anomalous non-Hermitian TAI phase characterized by a quantized polarization P_x \approx 0.25. Within this anomalous phase protected by the mobility gap, the zero-energy modes exhibit a distinct (N/2)-mode coalescence that scales with system size. Our findings demonstrate that non-Hermitian disorder engineered to preserve symmetry can induce and protect novel topological order inaccessible to conventional Hermitian disorder, thereby advancing the fundamental understanding of topological phenomena mediated by the interplay of disorder and non-Hermiticity. △ Less

Submitted 9 February, 2026; v1 submitted 1 February, 2026; originally announced February 2026.

Journal ref: Physical Review B 113, 064202 (2026)