Showing 1–50 of 452 results for author: Sinha, A

Scaling the Lightning Network with Practical Set Reconciliation

Authors: Xingyu Chen, Anish Sinha, David Starobinski, Ari Trachtenberg

Abstract: The Lightning Network (LN) utilizes gossip to share network topology, channel announcements and updates, and node announcements among its local constituents. Yet, our measurements show that this flooding-based gossip reconciliation is fundamentally inefficient. We propose, instead, to use set reconciliation protocols for sharing this information, and we systematically evaluate existing approaches… ▽ More The Lightning Network (LN) utilizes gossip to share network topology, channel announcements and updates, and node announcements among its local constituents. Yet, our measurements show that this flooding-based gossip reconciliation is fundamentally inefficient. We propose, instead, to use set reconciliation protocols for sharing this information, and we systematically evaluate existing approaches under realistic network conditions. We further propose ADAPTIVEIBLT, a novel adaptive IBLT (Invertible Bloom Lookup Table) protocol with a partial-decoding enhancement. By simulating reconciliation in Core-Lightning and evaluating real gossip snapshots, we demonstrate the practical benefits of reconciliation in scaling gossip reconciliation from hours down to a few minutes. △ Less

Submitted 16 August, 2026; originally announced August 2026.

Comments: Published in the 2026 IEEE International Conference on Blockchain and Cryptocurrency (ICBC 2026)

Journal ref: 2026 IEEE International Conference on Blockchain and Cryptocurrency (ICBC), 2026, pp. 1-5

Sequential Multimodal Evidence Optimization for Product Media Ranking in E-Commerce

Authors: Prasenjit Dey, Frank McIntyre, Arnab Sinha

Abstract: On modern e-commerce stores, customers consume ordered slates of heterogeneous product media, such as images, videos, and 3D renders, before making purchase decisions. Existing media-ranking systems often optimize myopic engagement proxies such as clicks or dwell time, even though product media assets are cooperative informational components of the same item that together help customers find the i… ▽ More On modern e-commerce stores, customers consume ordered slates of heterogeneous product media, such as images, videos, and 3D renders, before making purchase decisions. Existing media-ranking systems often optimize myopic engagement proxies such as clicks or dwell time, even though product media assets are cooperative informational components of the same item that together help customers find the information they need through sequential interaction. We present Sequential Multimodal Evidence Optimization (SMEO), a two-stage utility-guided framework for customer-oriented media sequencing. SMEO first learns a trajectory utility model from consumed media prefixes to estimate how ordered evidence helps customers reach a purchase decision, while mitigating position-bias and variable-depth imbalance in logged data. Recognizing that customer attention is a limited resource, it then trains an autoregressive ranking policy with survival-weighted reward-to-go that prioritizes the most decision-relevant information early, so customers can find what they need with less effort. By decoupling utility learning from policy optimization, SMEO enables stable offline learning from biased logs and post-hoc media attribution without explicit media-level labels. Evaluated offline on large-scale e-commerce sessions using doubly robust off-policy estimation, SMEO improves estimated conversion by 5.5% and helps customers reach a purchase decision with 15% fewer swipes than existing baselines. △ Less

Submitted 16 August, 2026; originally announced August 2026.

Comments: Proceedings of the 35th ACM International Conference on Information and Knowledge Management (CIKM 2026), Rome, Italy

Admissibility-Preserving Control for Strict-Feedback Nonlinear Systems with Asymmetric Actuator Constraints

Authors: Saurabh Kumar, Shashi Ranjan Kumar, Abhinav Sinha

Abstract: This paper develops Admissibility-Preserving Control (APC), a realization-centered safety-critical control framework for strict-feedback systems subject to asymmetric actuator limits, time-varying output constraints, and actuator-rate limitations. APC denotes the overall control architecture, whereas an Admissibility-Preserving Input Realization (APIR) denotes its constraint-realization module. Th… ▽ More This paper develops Admissibility-Preserving Control (APC), a realization-centered safety-critical control framework for strict-feedback systems subject to asymmetric actuator limits, time-varying output constraints, and actuator-rate limitations. APC denotes the overall control architecture, whereas an Admissibility-Preserving Input Realization (APIR) denotes its constraint-realization module. Therein, the APIR dynamically generates the physical plant input while rendering its prescribed asymmetric actuator set forward invariant. In contrast to algebraic clipping and post-design saturation compensation, the actuator limits are embedded directly in a continuously differentiable dynamic realization with user-selectable regularity and interpretable tuning parameters. The APIR is integrated with recursive backstepping by treating the realized plant input as an additional state. The resulting design does not require an input-to-state stability assumption on the uncontrolled plant. Instead, the nonlinear drift terms are compensated recursively, subject to an explicit compatibility condition between the desired motion, the available control authority, and the APIR interior gain. The framework is further extended to time-varying output-safe tracking through a smooth asymmetric logarithmic barrier coordinate and its associated Lyapunov function and to simultaneous actuator-magnitude and rate constraints through a cascaded APIR. Rigorous Lyapunov and invariance analyses establish regional asymptotic tracking, forward invariance of the compatible admissible sets, and boundedness of all closed-loop signals. Numerical studies illustrate asymmetric actuator utilization, output-safety preservation, and magnitude-rate constraint enforcement. △ Less

Submitted 15 August, 2026; originally announced August 2026.

LP-NAS: Linear Programming-based Neural Architecture Search

Authors: Abhishek Shukla, Ankur Sinha, Faiz Hamid

Abstract: Neural Architecture Search (NAS) aims to automate neural network architecture design, reducing reliance on human expertise. Among the various NAS methods, differentiable NAS has gained prominence due to its efficiency and accuracy compared to conventional NAS approaches. Since differentiable NAS relaxes the architecture search space into a continuous domain, it is possible to apply principles from… ▽ More Neural Architecture Search (NAS) aims to automate neural network architecture design, reducing reliance on human expertise. Among the various NAS methods, differentiable NAS has gained prominence due to its efficiency and accuracy compared to conventional NAS approaches. Since differentiable NAS relaxes the architecture search space into a continuous domain, it is possible to apply principles from continuous optimization to NAS. In this paper, we propose Linear Programming-based NAS (LP-NAS), a mathematical programming-based framework for differentiable NAS that is applicable to a wide range of continuous search spaces. LP-NAS formulates a linear program (LP) using the validation-loss gradient and the training-loss Hessian to compute an architecture update direction that improves generalization while preserving the optimality of the model parameters. By following this LP-derived descent direction, LP-NAS efficiently navigates the architecture search space, leading to faster and more effective architecture optimization. We introduce two computationally efficient variants of LP-NAS, namely S-LP-NAS and R-LP-NAS. Applying LP-NAS to the Differentiable Architecture Search (DARTS) search space results in two algorithmic variants, S-LP-DARTS and R-LP-DARTS. Both variants achieve faster convergence and significantly higher validation performance during the early search iterations than the standard DARTS algorithm. Extensive experiments on CIFAR-10 and CIFAR-100 show that LP-DARTS outperforms standard DARTS in both the architecture search and evaluation phases. Additionally, we compare our approach with several DARTS variants (P-DARTS, PC-DARTS, and STO-DARTS) on the CIFAR-10 dataset and demonstrate its effectiveness. Furthermore, we validate the transferability of the discovered architectures through experiments on the ImageNet dataset. △ Less

Submitted 14 August, 2026; originally announced August 2026.

Comments: 20 pages, 5 figures

Designing Compact Neural Architectures via Neuron Gating and Mixed Activation

Authors: Abhishek Shukla, Ankur Sinha, Faiz Hamid

Abstract: Neural Architecture Search (NAS) is naturally formulated as a bilevel optimization problem, where the upper-level optimizes the architecture using validation performance and the lower-level trains network parameters using training loss. However, NAS is computationally expensive due to discrete architectural decisions, exponentially growing search spaces, and the high cost of training candidate arc… ▽ More Neural Architecture Search (NAS) is naturally formulated as a bilevel optimization problem, where the upper-level optimizes the architecture using validation performance and the lower-level trains network parameters using training loss. However, NAS is computationally expensive due to discrete architectural decisions, exponentially growing search spaces, and the high cost of training candidate architectures. This work develops a general bilevel optimization framework for NAS across diverse architectures, including MLPs, CNNs, RNNs, and Transformers, to identify compact architectures with strong predictive performance. We propose three scalable formulations that replace discrete neuron- and activation-level decisions with continuous relaxations, enabling differentiable optimization over otherwise combinatorial architecture spaces. These formulations give rise to three NAS methods: NAS based on Neuron Gating (NAS-NG), NAS based on Mixed Activation (NAS-MA), and NAS based on Neuron Gating and Mixed Activation (NAS-NGMA). Experiments on MLPs and CNNs using MNIST and CIFAR-10 show that the proposed methods consistently identify compact architectures with competitive or improved predictive performance. On MNIST, NAS-NGMA achieves 98.68% test accuracy with 7.69M MLP parameters, while NAS-NG achieves 99.63% accuracy with only 0.26M CNN parameters. On CIFAR-10, the proposed methods consistently outperform vanilla DARTS. Further experiments demonstrate that NAS-NG can optimize substantially over-parameterized and literature-optimal architectures, improving accuracy while reducing parameters. These results establish relaxed bilevel optimization as a scalable alternative to discrete NAS and provide a general framework for efficient neuron- and activation-level architecture optimization. △ Less

Submitted 14 August, 2026; originally announced August 2026.

Comments: 33 pages, 17 figures

Can Humans Dream of Electric Sheep? Human-Written Samples for Fine-Grained Vision-and-Language Hallucination Benchmarking

Authors: Timothee Mickus, Claudio Savelli, Eduardo Calò, Emilio Raimond, Stella Frank, Hengyu Luo, Flavio Giobergia, Vincent Segonne, Chuyuan Li, Aman Sinha, Lorenzo Vaiani, Jörg Tiedemann, Raúl Vázquez

Abstract: In an age of rapid model turnover, how do we make hallucination evaluation more perennial? We explore whether human-written hallucination samples could take the place of model-generated hallucinations, in order to make benchmarking detection independent of particular models. To this end, we construct a dataset of 1,600 human-written samples, spanning four languages (Chinese, English, French, Itali… ▽ More In an age of rapid model turnover, how do we make hallucination evaluation more perennial? We explore whether human-written hallucination samples could take the place of model-generated hallucinations, in order to make benchmarking detection independent of particular models. To this end, we construct a dataset of 1,600 human-written samples, spanning four languages (Chinese, English, French, Italian), and 18,400 samples from five vision-and-language models, all annotated for hallucinations using a fine-grained span-level labeling scheme. We find that human-written samples result in higher agreement and allow greater control of dataset contents, while remaining distributionally similar to samples derived from vision-and-language samples and providing a reasonable portrayal of detection capabilities - suggesting that human data is a viable substitute for model-based hallucination benchmarks. △ Less

Submitted 2 August, 2026; originally announced August 2026.

BoilerSketch: A TA-Supervised, Diagram-First GenAI Practice for Structured Diagrams in CS1/Early CS2

Authors: Ethan Dickey, Vivan Tiwari, Anvit Sinha, Andres Bejarano

Abstract: This innovative practice full paper presents BoilerSketch, a TA-supervised, diagram-first GenAI practice and tablet interface for providing structured visual explanations in CS1 and early CS2 support settings. Large early computing courses routinely face a support bottleneck during labs and office hours because many student questions are best answered with a diagram rather than additional text, ye… ▽ More This innovative practice full paper presents BoilerSketch, a TA-supervised, diagram-first GenAI practice and tablet interface for providing structured visual explanations in CS1 and early CS2 support settings. Large early computing courses routinely face a support bottleneck during labs and office hours because many student questions are best answered with a diagram rather than additional text, yet most AI tutoring tools remain text-forward and unreliable at producing accurate, pedagogically useful visuals. BoilerSketch addresses this gap through a dual-pane interaction model that combines chat with a pen-enabled whiteboard for student sketches and a prompting strategy that constrains the model to generate structured, renderable Mermaid diagrams rather than free-form images. To preserve academic integrity, the system is intentionally scoped to conceptual explanation: it forbids executable code and code-level debugging and uses a human-in-the-loop workflow in which teaching assistants remain accountable supervisors who can monitor sessions and intervene when responses require correction, deeper probing, or escalation to live help. We report a 45-minute expert evaluation with 21 instructional staff from a large programming course who used BoilerSketch on representative questions and completed a post-use survey. Two-thirds rated the system at least moderately helpful for conceptual understanding and at least moderately useful for typical support tasks. Staff saw the strongest value in routine diagram-based explanations and noted limits in diagram depth and applicability to more advanced topics. We conclude with practical guidance for adopting supervised, diagram-first GenAI support in early computing courses, emphasizing scope-and-escalation rules, prompt-as-policy guardrails, and reliable structured diagram rendering. △ Less

Submitted 1 August, 2026; originally announced August 2026.

Comments: 9 pages, 3 tables, 3 figures

ACM Class: K.3.2; K.3.1; H.5.2

PTP: Previous-Token Prediction based LLM Inversion for Near-Exact Prompt Reconstruction

Authors: Pirzada Suhail, Nagasai Saketh Naidu, Atanu R Sinha, Amit Sethi

Abstract: Large language models (LLMs) generate text by auto-regressively sampling the next token. This inherently leads to a many-to-many mapping between prompts and responses, complicating the task of inferring prompts from observed outputs. Prior work on LLM inversion frames prompt recovery as a semantic reconstruction task. They rely on fine-tuning pretrained sequence-to-sequence models on large externa… ▽ More Large language models (LLMs) generate text by auto-regressively sampling the next token. This inherently leads to a many-to-many mapping between prompts and responses, complicating the task of inferring prompts from observed outputs. Prior work on LLM inversion frames prompt recovery as a semantic reconstruction task. They rely on fine-tuning pretrained sequence-to-sequence models on large external datasets--and requiring access to model weights or logits--to generate semantically plausible prompts. In contrast, we present a functional approach to inverting a given LLM in a black-box setting, without auxiliary aids. We train an explicit inverse language model entirely from scratch on data synthetically generated from the target LLM itself. Analogous to forward next-token prediction, our inverse model is trained using previous-token prediction, establishing a generative link between the forward and inverse processes that enables faithful prompt reconstruction. Moreover, it naturally supports diverse prompt reconstructions through sampling, whereby all such prompts induce similar responses under the forward, target LLM. Our approach generalises across datasets and exhibits transferability in reconstructing prompts from responses generated by different LLMs. Further, across the set of token based evaluation metrics for prompt and response reconstructions, our approach outperforms prior work. △ Less

Submitted 31 July, 2026; originally announced July 2026.

Baikal: Structured Search for Deep Research over Data Lakes

Authors: Dhruv Agarwal, Rishitha Guttapalle Mohan, Aarti Kumari, Ashi Sinha, Athulya Anil, Kavitha Srinivas, Horst Samulowitz, Andrew McCallum

Abstract: Deep research over data lakes requires an LLM agent to investigate evidence across thousands of heterogeneous tables and passages to synthesize a report. Existing methods perform iterative retrieval and generation, letting accumulated context determine what to investigate next, which can overexploit locally promising evidence and fail to cover distinct semantic regions under a fixed budget. To add… ▽ More Deep research over data lakes requires an LLM agent to investigate evidence across thousands of heterogeneous tables and passages to synthesize a report. Existing methods perform iterative retrieval and generation, letting accumulated context determine what to investigate next, which can overexploit locally promising evidence and fail to cover distinct semantic regions under a fixed budget. To address this, we cast deep research over data lakes as a budgeted search problem and present Baikal - a framework that clusters heterogeneous evidence into semantic regions, then searches over them adaptively to balance exploration and exploitation. Within each selected region, Baikal generates and investigates region-grounded subquestions, using finding quality as rewards to update region-level value estimates and guide search under policies ranging from random and LLM-guided selection to Bayesian $ε$-greedy and UCB. We evaluate Baikal on 15 queries each over HybridQA and TAT-QA data lakes containing 10,993 and 2,757 tables, respectively, together with 227K Wikipedia passages and 13K financial report passages. We assess research quality with a new rubric covering groundedness, relevance, diversity, and utility, and use GPT-5-mini to score Baikal and strong baselines, including DeepSearcher and an OpenCode research agent with retrieval and clustering variants. Across both data lakes, Baikal performs strongly under several region-selection policies; its best configuration improves report scores over the strongest baselines by 28% on HybridQA and 36% on TAT-QA. Our analyses attribute these gains to organizing and exploring semantic evidence regions, which improves groundedness and diversity and yields more useful findings under the same subquestion budget. These results demonstrate the value of structured semantic exploration for systematic research and discovery over heterogeneous data lakes. △ Less

Submitted 30 July, 2026; originally announced July 2026.

Trainable Nonexpansive Denoisers for Contractive Image Reconstruction

Authors: Arghya Sinha, Aditya Banerjee, Trishit Mukherjee, Kunal N. Chaudhury

Abstract: Trainable denoisers with Lipschitz control have become central to convergent image reconstruction. However, training neural networks that simultaneously offer strong denoising performance and global Lipschitz guarantees is challenging. Existing approaches enforce Lipschitz control only empirically, providing no guarantees beyond the training data. In this work, we show that by exploiting the actio… ▽ More Trainable denoisers with Lipschitz control have become central to convergent image reconstruction. However, training neural networks that simultaneously offer strong denoising performance and global Lipschitz guarantees is challenging. Existing approaches enforce Lipschitz control only empirically, providing no guarantees beyond the training data. In this work, we show that by exploiting the action of permutations on the image lattice, we can constrain a neural architecture that is globally nonexpansive (Lipschitz bound $\leqslant 1$). We integrate the proposed denoiser with forward imaging operators to develop a reconstruction mechanism that is provably contractive and therefore globally convergent. Experiments on standard inverse problems, such as superresolution and deblurring, demonstrate that our reconstruction performance is competitive with softly constrained baselines while providing Lipschitz guarantees. △ Less

Submitted 25 July, 2026; originally announced July 2026.

Comments: accepted at ICML 2026

Stabilizing Deep Reconstruction Operators with Contractive Anchoring

Authors: Arghya Sinha, Trishit Mukherjee, Kunal N. Chaudhury

Abstract: Pretrained deep denoisers can be used to solve a wide range of model-based image reconstruction tasks via Plug-and-Play (PnP) and Regularization-by-Denoising (RED) algorithms, without retraining per task. These denoisers are trained only for single-step denoising. Using them as Image Reconstruction (IR) regularizers in an iterative process can destabilize reconstruction. A common failure mode is t… ▽ More Pretrained deep denoisers can be used to solve a wide range of model-based image reconstruction tasks via Plug-and-Play (PnP) and Regularization-by-Denoising (RED) algorithms, without retraining per task. These denoisers are trained only for single-step denoising. Using them as Image Reconstruction (IR) regularizers in an iterative process can destabilize reconstruction. A common failure mode is the peak-and-collapse behaviour: metrics such as PSNR improve for early iterations and then abruptly degrade, making these algorithms unreliable in practice. We propose a data-driven stabilization framework that (i) formalizes this instability of any IR operator through a local quantity and (ii) prevents collapse by regularizing this quantity adaptively, requiring no retraining or modification of the given pretrained network. Our key idea is to control the potentially unstable IR operator with a contractive operator whose stable iterates act as an anchor and prevent collapse. We further introduce an efficient family of trainable contractive operators that serve as strong anchors while remaining lightweight. Extensive experiments across proximal algorithms, denoiser architectures, noise levels, and imaging tasks show consistent, collapse-free performance and improved reliability of PnP and RED reconstruction. △ Less

Submitted 29 July, 2026; v1 submitted 25 July, 2026; originally announced July 2026.

Comments: Accepted at ECCV 2026

MSC Class: 68U10; 94A08; 47H09; 47J25

Robust Critics: Defending LLMs Against Multi-Turn Attacks

Authors: Roman Belaire, Arunesh Sinha, Pradeep Varakantham

Abstract: When a user asks a language model something harmful, is it a genuine attack or a misunderstood but well-meaning question? This ambiguity is one of the central challenges of LLM safety. A model that assumes the worst harms legitimate users; one that assumes the best is easily exploited. The problem is compounded in multi-turn dialogue, where an attacker's true intent may only reveal itself graduall… ▽ More When a user asks a language model something harmful, is it a genuine attack or a misunderstood but well-meaning question? This ambiguity is one of the central challenges of LLM safety. A model that assumes the worst harms legitimate users; one that assumes the best is easily exploited. The problem is compounded in multi-turn dialogue, where an attacker's true intent may only reveal itself gradually across many exchanges, yet existing safety frameworks apply a contextual bandit treatment, ignoring the trajectory of the conversation. To that end, we propose Dialogue Critic Guided Sampling (DCGS), a framework that addresses this by inferring user intent at every turn of dialogue. Instead of applying a fixed rule about what is or is not safe, DCGS learns what the user's intent is likely to be based on the full conversational history and generates responses accordingly. Formally, we model adversarial dialogue as a Markov Decision Process and learn value and regret-based critics at both the individual token and utterance (full response) levels, scoring candidate responses via an action-value critic. We prove that this inference-time reweighting approximates exponential tilting of the base policy, guaranteeing improvement in expected return for any finite candidate pool, a property that group-relative objectives do not exhibit. Evaluated on CARES-18k, WildJailbreak, Redbench, and Harmbench, DCGS outperforms strong robust baselines and frontier models on adversarial dialogue tasks. DCGS also transfers to frontier models, improving their robustness without fine-tuning. △ Less

Submitted 24 May, 2026; originally announced July 2026.

Contact-Persistent Full Actuation for Aerial Physical Interaction

Authors: Abhimanyu Khadga, Abhinav Sinha, Shashi Ranjan Kumar

Abstract: Fully actuated unmanned aerial vehicles (UAVs) are usually certified through rank conditions on a control-allocation matrix or through free-flight tracking performance. For aerial physical interaction, this certification may be incomplete. During sustained contact, part of the available wrench is consumed by the interaction task, and only the residual wrench remains available for stabilization, di… ▽ More Fully actuated unmanned aerial vehicles (UAVs) are usually certified through rank conditions on a control-allocation matrix or through free-flight tracking performance. For aerial physical interaction, this certification may be incomplete. During sustained contact, part of the available wrench is consumed by the interaction task, and only the residual wrench remains available for stabilization, disturbance rejection, and maneuvering. This paper introduces a control-theoretic framework for \emph{contact-persistent full actuation}. A rigid-body model on $\R^{3}\times\SO\left(3\right)$ is combined with a morphology-dependent wrench map that captures fixed-tilt, variable-tilt, coaxial, and overactuated multirotor architectures. We define feasible wrench sets under actuator limits, residual wrench sets under task loading, and residual authority margins that strengthen the usual rank-based notion of full actuation. The main result shows that contact-persistent full actuation is equivalent to interiority of the task wrench in the constrained feasible wrench polytope, and that the residual authority radius is exactly the distance to the polytope boundary. We further introduce a signed residual-margin certificate for infeasible and boundary cases, a slack-maximizing allocation certificate, and a robust implementability condition that can be used as a margin-aware safety filter. Numerical evaluation on an abstract tilted hexarotor shows that full row rank alone does not imply feasible contact operation. Intermediate tilt angles preserve residual authority during pushing, whereas small or excessive tilts fail because of lateral-force deficiency or hover-margin loss. △ Less

Submitted 21 July, 2026; originally announced July 2026.

A Formal Hierarchical Architecture for Agentic Orchestration with Stack-Based Execution and Lazy Discovery

Authors: Prashant Devadiga, Abhishek, Adithya Mishra, Alok Singh, Amisha Sinha, Asit Desai, Gaurang Dahad, Harshit Bhushan, Mandati Pramod Reddy, Prakhar Gupta, Rupesh Patil, Siddhi Behere

Abstract: The rapid expansion of capabilities in Large Language Model (LLM) agents has exposed a critical architectural bottleneck: when agents are given access to a flat, monolithic registry of tools, the model must evaluate hundreds or thousands of options simultaneously. This leads to decision-space explosion, context window saturation, and degraded routing accuracy. To address these limitations, this pa… ▽ More The rapid expansion of capabilities in Large Language Model (LLM) agents has exposed a critical architectural bottleneck: when agents are given access to a flat, monolithic registry of tools, the model must evaluate hundreds or thousands of options simultaneously. This leads to decision-space explosion, context window saturation, and degraded routing accuracy. To address these limitations, this paper presents a hierarchical, skill-based architecture for agentic orchestration. Capabilities are organized as a rooted tree where internal nodes make routing decisions and leaf nodes execute deterministic tasks. The runtime enforces a single-step execution loop governed by a Last-In-First-Out (LIFO) stack, giving the agent a form of memory akin to a Pushdown Automaton, therefore enabling it to track nested execution contexts and resume deterministically from any depth. Capability discovery follows a manifest-driven, lazy-loading protocol: only the immediate children of the active node are loaded, so memory and prompt costs scale with the explored path rather than the global registry. By replacing global memory with localized stack frames, the architecture prevents outputs from one execution branch from leaking into another, establishing the isolation guarantees required for deployment in regulated enterprise environments. We also discuss UPI Help, an AI-powered digital payments support product, as a motivating production deployment context. We provide a mathematical formalization of the orchestration state, detailed algorithmic analysis of the execution loop, and controlled benchmarks comparing flat and hierarchical routing under increasing tool catalogs, multi-step workflow pressure, and visible schema-token exposure per LLM call. △ Less

Submitted 13 July, 2026; originally announced July 2026.

Flowcode: An AI-Powered Programming Environment for Scaffolding Iteration in Creative Computing Education

Authors: Tiffany Tseng, Liliana Hanem Seoror, Jeevika Adda, Meitalia Factor, Rona Darabi, Kiley R Matschke, Tiffany Fu, Annie Lin, Alekhya Maram, Arya Sinha

Abstract: Building upon found examples is a popular way people learn to code, especially in creative coding communities where sharing projects and remixing are common practices. But effectively doing so requires being able to 1) understand how existing code works, and 2) extend it by writing code that implements your own ideas, practices that can be challenging for new creative coders. We explored how to su… ▽ More Building upon found examples is a popular way people learn to code, especially in creative coding communities where sharing projects and remixing are common practices. But effectively doing so requires being able to 1) understand how existing code works, and 2) extend it by writing code that implements your own ideas, practices that can be challenging for new creative coders. We explored how to support these two processes through the design of Flowcode, a creative coding programming environment that integrates a flowchart for visualizing code structure and a chat interface tailored to support learning to code over vibe coding. We share how we iterated on the design of Flowcode over two studies with new creative coders, reflecting on the roles visualization and friction may play in enabling productive AI-use in computing education. △ Less

Submitted 7 July, 2026; originally announced July 2026.

A Step Towards Robust Unsupervised Domain Adaptation via Fine-Tuning and Reinforcement Learning

Authors: Sushant Dagaji Desale, Rahul Mishra, Ashutosh Kumar Sinha

Abstract: Adversarial robustness in Unsupervised Domain Adaptation (UDA) remains a significant challenge due to noisy pseudo labels and inherent distributional shifts between the clean source and adversarially perturbed target domains. Existing approaches often fail to achieve an optimal trade-off between robustness and accuracy, as pseudo-labels generated by domain-adapted models tend to introduce classifi… ▽ More Adversarial robustness in Unsupervised Domain Adaptation (UDA) remains a significant challenge due to noisy pseudo labels and inherent distributional shifts between the clean source and adversarially perturbed target domains. Existing approaches often fail to achieve an optimal trade-off between robustness and accuracy, as pseudo-labels generated by domain-adapted models tend to introduce classification errors under adversarial attacks. In this work, we propose \textbf{SFT+RL}, a two-stage robust UDA framework that integrates Supervised Fine Tuning (SFT) and Reinforcement Learning (RL) on top of CLIP's pre-trained visual encoder. In the SFT stage, we adversarially fine-tune a linear classifier using PGD-based perturbations over the labelled source domain while partially unfreezing CLIP's projection layer. It allows adaptation to adversarial noise while preserving CLIP's rich semantic priors. We introduce a confidence-guided pseudo-labeling strategy in the RL stage to annotate unlabeled target samples progressively. Pseudo labels are filtered using a decaying confidence threshold to balance quality and coverage, and the model is trained on a composite dataset formed by combining clean source samples with high-confidence target samples. Adversarial training is applied to mixed batches of clean and adversarial examples to enhance cross-domain robustness. Comprehensive evaluations on three benchmark datasets OfficeHome~\cite{tomm-ude}, PACS~\cite{pacs}, and VisDA~\cite{visda} demonstrate the effectiveness of our approach. Notably, \textbf{SFT+RL} achieves average improvements of \textbf{10.2\%} in clean accuracy and \textbf{15.8\%} in adversarial robustness across all three datasets, outperforming existing state-of-the-art methods. △ Less

Submitted 3 July, 2026; originally announced July 2026.

Comments: Accepted and presented at the 28th European Conference on Artificial Intelligence (ECAI 2025), Bologna, Italy

Journal ref: Frontiers in Artificial Intelligence and Applications, Volume 413, ECAI 2025, IOS Press, 2025

AI usage patterns are shaped by perceived gains in human agency

Authors: Ian Beacock, Rachel Xu, Laura Murray, Patrick Anson, Beth Goldberg, Devika Kumar, Jun Lee, Rebekah Park, Anoop Sinha

Abstract: As conversational AI systems become more deeply integrated into daily life, the implications for human agency are increasingly urgent to understand. AI's potential to amplify capability sits alongside risks of individual and collective disempowerment, yet empirical, ecologically-valid evidence about cumulative usage is scarce. We analyze deep ethnographic data from a study of daily AI chatbot user… ▽ More As conversational AI systems become more deeply integrated into daily life, the implications for human agency are increasingly urgent to understand. AI's potential to amplify capability sits alongside risks of individual and collective disempowerment, yet empirical, ecologically-valid evidence about cumulative usage is scarce. We analyze deep ethnographic data from a study of daily AI chatbot users (n = 51) in the United States, Germany, and Singapore to illuminate conversational AI usage in situated context as a sociotechnical practice. We show that people consistently link sustained AI usage to perceived gains in individual agency. Crucially, these perceived gains often outweigh concerns about accuracy, reliability, and consistency to shape usage patterns. Our findings challenge prevailing assumptions about how and why humans use AI systems over time, suggesting that traditional trust-based models are not sufficient for explaining human behavior with conversational AI. Finally, we expose a critical tension: immediate psychological boosts to perceived agency may not necessarily translate into material effects, structural empowerment, or long-term capacity. Our results help establish a new foundation for novel behavioral frameworks, measurement tools, and AI benchmarks to ensure conversational AI strengthens human agency in substantial, sustained ways. △ Less

Submitted 2 July, 2026; originally announced July 2026.

Bilevel Optimization for Neural Architecture Search

Authors: Abhishek Shukla, Ankur Sinha, Faiz Hamid

Abstract: Bilevel optimization has become an influential and widely adopted framework for addressing hierarchical optimization problems in machine learning, providing an effective approach to modeling the interaction between two levels of optimization, with applications such as hyperparameter tuning, meta-learning, adversarial training, and data poisoning. Neural Architecture Search (NAS), a subfield of hyp… ▽ More Bilevel optimization has become an influential and widely adopted framework for addressing hierarchical optimization problems in machine learning, providing an effective approach to modeling the interaction between two levels of optimization, with applications such as hyperparameter tuning, meta-learning, adversarial training, and data poisoning. Neural Architecture Search (NAS), a subfield of hyperparameter optimization, is a prime example of a bilevel optimization problem, with architecture parameters optimized at the outer-level and network weights optimized at the inner level. This paper presents a structured overview of NAS through the lens of bilevel optimization. We categorize existing NAS approaches into two main classes: sampling-based methods, which search optimal architectures using different architecture samplers, and bilevel theory-based methods, which solve the architecture search problem using bilevel optimization principles. We further highlight our current research direction, wherein the bilevel NAS formulation is addressed through an auxiliary mathematical programming framework. This framework enables the systematic integration of second-order information from the model's training loss function and ensures the optimality of the model parameters while modifying architecture parameters. By simultaneously updating the architecture and model parameters along their respective optimal descent directions derived from the auxiliary mathematical program, these methods achieve more principled and theoretically consistent results. The same auxiliary program can also be used for simultaneous hyperparameter and model fine-tuning. A comparative analysis shows that bilevel theory-based approaches generally outperform sampling-based methods, both in accuracy and efficiency. △ Less

Submitted 28 June, 2026; originally announced June 2026.

Comments: 48 pages, 20 figures

How Well Do Self-Supervised Speech Models Encode Age and Gender in Children's Speech? A Layer-Wise Analysis Across Multiple Architectures

Authors: Abhijit Sinha, Hemant Kumar Kathania, Mohit Joshi, Harishankar Kumar, Shrikanth Narayanan, Sudarsana Reddy Kadiri

Abstract: Self-supervised learning (SSL) models have become a central component of modern speech processing systems, as they enable the learning of rich acoustic representations without reliance on labeled data. Despite their success on adult speech, it remains unclear how effectively these models capture speaker-related attributes such as age and gender in children's speech, which differs substantially fro… ▽ More Self-supervised learning (SSL) models have become a central component of modern speech processing systems, as they enable the learning of rich acoustic representations without reliance on labeled data. Despite their success on adult speech, it remains unclear how effectively these models capture speaker-related attributes such as age and gender in children's speech, which differs substantially from adult speech due to ongoing physiological and cognitive development. Higher pitch, increased articulatory variability, and age-dependent acoustic changes make children's speech a particularly challenging domain. In this work, we present a comprehensive analysis of how age and gender information is encoded across layers of four widely used SSL models: Wav2Vec2, HuBERT, Data2Vec, and WavLM. Layer-wise features are extracted and evaluated using a lightweight CNN on two benchmark children's speech corpora, PFSTAR and CMU Kids. To analyze feature compactness and redundancy, PCA is applied to identify redundancy and highlight the dimensions that contribute most to classification performance. Experimental results show that age- and gender-related information is unevenly distributed across SSL layers, with early to mid-level layers encoding the strongest paralinguistic cues. HuBERT achieves the best overall performance for age classification, while Wav2Vec2 and HuBERT lead gender classification on PFSTAR and CMU Kids, respectively. Beyond single-split evaluation, we further demonstrate that these findings remain stable under speaker-wise cross-validation, layer aggregation, and cross-database evaluation, indicating robustness to data imbalance and domain mismatch. Finally, we show that reliable age and gender classification is achievable even from short speech segments of 1--3 seconds. △ Less

Submitted 20 June, 2026; originally announced June 2026.

TimeProVe: Propose, then Verify for Efficient Long Video Temporal Reasoning in Activities of Daily Living

Authors: Arkaprava Sinha, Dominick Reilly, Siddharth Krishnan, Hieu Le, Srijan Das

Abstract: Long Video Question Answering (LVQA) requires identifying sparse, query-relevant evidence within hours-long untrimmed videos. Existing approaches either process videos densely with large vision-language models (VLMs), incurring prohibitive computational cost, or rely on sparse caption-based reasoning, which often misses temporally localized and motion-centric evidence. We introduce TimeProVe, a co… ▽ More Long Video Question Answering (LVQA) requires identifying sparse, query-relevant evidence within hours-long untrimmed videos. Existing approaches either process videos densely with large vision-language models (VLMs), incurring prohibitive computational cost, or rely on sparse caption-based reasoning, which often misses temporally localized and motion-centric evidence. We introduce TimeProVe, a cost-efficient hybrid framework for temporally grounded reasoning in long videos. TimeProVe first employs lightweight modules to generate action-grounded answer--evidence hypotheses and subsequently invokes an expensive VLM only for targeted verification. The core of our framework lies in the Action-based Candidate Evidence (ACE) module, which converts temporally localized actions into query-conditioned candidate answers and supporting evidence windows through lightweight LLM reasoning. We further introduce OpenTSUBench (OTB), an open-ended benchmark designed to evaluate temporally grounded reasoning in real-world Activities of Daily Living (ADL) scenarios. Experiments show that TimeProVe outperforms the strongest baseline on OTB by 7.3%, while reducing VLM calls by 75% and inference cost by 93%. Furthermore, without explicit temporal grounding training, TimeProVe achieves competitive performance on Charades-STA, and reaches state-of-the-art results when enhanced with grounding VLMs. △ Less

Submitted 18 June, 2026; originally announced June 2026.

UNIEGO: Proxies as Mediators for Unified Egocentric Video Representation Learning

Authors: Wenhao Chi, Arkaprava Sinha, Dominick Reilly, Hieu Le, Srijan Das

Abstract: Egocentric video understanding is inherently limited by the narrow perspective of wearable cameras: a single viewpoint, a single modality, a single model cannot capture the full richness of human action. We argue that a truly expressive egocentric representation must subsume complementary knowledge across viewpoints, modalities, and foundation model representations, yet remain deployable from egoc… ▽ More Egocentric video understanding is inherently limited by the narrow perspective of wearable cameras: a single viewpoint, a single modality, a single model cannot capture the full richness of human action. We argue that a truly expressive egocentric representation must subsume complementary knowledge across viewpoints, modalities, and foundation model representations, yet remain deployable from egocentric video alone. To this end, we introduce a hierarchical multi-teacher distillation framework that produces UNIEGO, a unified egocentric encoder trained with nine teachers spanning ego-exo viewpoints, RGB, depth, and skeleton modalities, and four foundation models. Rather than distilling directly from heterogeneous teachers whose incompatible architectures and feature geometries induce conflicting gradients, our framework interposes a layer of representation-specific Proxy models that translate diverse teacher knowledge into a homogeneous egocentric space. A second distillation stage, Selective Proxy Distillation (SPD), then adaptively selects, for each training sample, the subset of proxies that are both correct and confident, distilling exclusively from reliable supervision and suppressing erroneous signals. SPD is further stabilized by initializing UNIEGO as a learned convex combination of proxy parameters, placing the unified model in a well-conditioned region of the loss landscape before distillation begins. UNIEGO achieves state-of-the-art performance across three egocentric video understanding tasks - action recognition, video retrieval, and action segmentation on three challenging ego-exo benchmarks, outperforming naive multi-teacher distillation baselines and demonstrating that structured, proxy-mediated knowledge transfer yields richer and more discriminative egocentric representations. △ Less

Submitted 18 June, 2026; originally announced June 2026.

Human-on-the-Loop Orchestration for AI-Assisted Legal Discovery

Authors: Anushree Sinha, Srivaths Ranganathan, Abhishek Dharmaratnakar, Debanshu Das

Abstract: Autonomous Large Language Model (LLM) agents are increasingly deployed in electronic discovery (e-discovery), where compounding errors across multi-step reasoning chains can constitute legal malpractice. Unlike single-turn retrieval, agentic workflows operating over privileged document corpora exhibit a class of failure we term "trajectory collapse": an early misclassification silently propagates,… ▽ More Autonomous Large Language Model (LLM) agents are increasingly deployed in electronic discovery (e-discovery), where compounding errors across multi-step reasoning chains can constitute legal malpractice. Unlike single-turn retrieval, agentic workflows operating over privileged document corpora exhibit a class of failure we term "trajectory collapse": an early misclassification silently propagates, rendering an entire privilege review invalid. This paper makes three contributions. First, we propose a structured taxonomy of agentic failures in legal information retrieval, organized by functional stage. Second, we introduce a four-layer verification architecture -- spanning planning, reasoning, execution, and uncertainty quantification -- designed to intercept these failures before they compound. Third, we present a preliminary simulation study on a synthetic e-discovery corpus that demonstrates how mandatory Human-on-the-Loop (HOTL) escalation thresholds reduce privilege-waiver risk relative to fully autonomous baselines. Our results suggest that calibrated uncertainty thresholds can reduce privilege-waiver risk by up to 61% versus fully autonomous deployment, while routing fewer than one quarter of documents to attorney review. △ Less

Submitted 18 June, 2026; originally announced June 2026.

Cross-Dataset, Age, and Gender Generalization: A Comprehensive Analysis of Fine-Tuning Strategies for Low-Resource Children's ASR

Authors: Abhijit Sinha, Hemant Kumar Kathania, Sudarsana Reddy Kadiri, Shrikanth Narayanan

Abstract: The challenge associated with recognizing dysarthric speech primarily arises from pronounced acoustic variability attributed to impaired articulatory precision. Past research has demonstrated improved recognition through the use of hybrid DNN/HMM sequence discriminative training. This paper presents a comprehensive investigation of various combinations of acoustic features tailored to different Ac… ▽ More The challenge associated with recognizing dysarthric speech primarily arises from pronounced acoustic variability attributed to impaired articulatory precision. Past research has demonstrated improved recognition through the use of hybrid DNN/HMM sequence discriminative training. This paper presents a comprehensive investigation of various combinations of acoustic features tailored to different Acoustic Models, offering suitable feature selections for each. The incorporation of Pitch features notably improved recognition performance, especially for sentence recognition tasks involving dysarthric speech. Through a systematic examination of the TORGO database, we have demonstrated the potential to enhance the performance of the state-of-the-art Factorized Time Delay Neural Network (F-TDNN) model for recognizing dysarthric speech. Our methods, implemented with the F-TDNN model, resulted in a 4.65\% relative improvement in isolated word recognition and a 4.63\% relative improvement in sentence recognition for dysarthric speech, compared to previous research. This improvement effectively compensates for speech variability, attributable to our deliberate selection of the number of overlapping frames between consecutive training example chunks. △ Less

Submitted 22 June, 2026; v1 submitted 18 June, 2026; originally announced June 2026.

LiFT: Local Search via Linear Programming for Overfitting-Controlled Transformers

Authors: Abhishek Shukla, Anikeit Khanna, Ankur Sinha, Faiz Hamid

Abstract: This paper proposes a Linear Programming (LP)-based local search framework for fine-tuning pretrained transformer models with explicit control against overfitting. The approach formulates transformer fine-tuning as a bilevel optimization-based regularization problem, in which model parameters and regularization hyperparameters are jointly updated. Information collected during initial warm-up itera… ▽ More This paper proposes a Linear Programming (LP)-based local search framework for fine-tuning pretrained transformer models with explicit control against overfitting. The approach formulates transformer fine-tuning as a bilevel optimization-based regularization problem, in which model parameters and regularization hyperparameters are jointly updated. Information collected during initial warm-up iterations, including validation gradients and training Hessian information, is used to construct a local descent direction by solving an LP that minimizes a scaled directional derivative while preserving training optimality. This validation-aware descent direction enables focused local updates of both parameters and regularization hyperparameters, reducing overfitting without requiring repeated full retraining cycles. The resulting method, termed Linear Programming-based Fine-Tuning (LiFT) for transformers, differs from conventional fine-tuning by systematically identifying task-specific updates rather than relying on heuristic or grid-based hyperparameter selection. Experiments on GPT-2 Small fine-tuned on WikiText-2 demonstrate that LiFT enables effective adaptation through selective tuning of transformer blocks and regularization parameters, yielding consistent improvements in test perplexity across multiple layer configurations and regularization settings, with particularly pronounced gains in overfitting-prone scenarios. Beyond empirical performance, LiFT establishes a principled connection between transformer fine-tuning, bilevel optimization, local search, and regularization theory. △ Less

Submitted 15 June, 2026; originally announced June 2026.

Comments: 22 pages, 6 figures, published in The 20th Learning and Intelligent Optimization Conference (LION 2026)

Distributed Safe Consensus Under Asymmetric Input and Time-Varying Output Constraints

Authors: Abhinav Sinha, Shashi Ranjan Kumar

Abstract: This paper studies safe distributed consensus for single-integrator multi-agent systems over connected undirected graphs under simultaneous asymmetric actuator constraints and output safety constraints. Each agent is equipped with a continuously differentiable asymmetric actuator dynamics that maps a commanded control signal to the realized plant input while keeping the latter strictly inside a pr… ▽ More This paper studies safe distributed consensus for single-integrator multi-agent systems over connected undirected graphs under simultaneous asymmetric actuator constraints and output safety constraints. Each agent is equipped with a continuously differentiable asymmetric actuator dynamics that maps a commanded control signal to the realized plant input while keeping the latter strictly inside a prescribed admissible interval. To address output safety, a barrier-coordinate transformation is introduced over a common time-varying safe interval, and a distributed synchronization law is designed in the transformed coordinates. The resulting controller integrates a graph-based coordination layer with an actuator-side tracking layer, thereby enabling simultaneous enforcement of input admissibility, forward invariance of the safe output set, and asymptotic synchronization. For compact admissible sets of initial conditions, it is shown that the closed-loop solution is complete, all signals remain bounded, the actuator inputs remain strictly within their asymmetric bounds, and the agent outputs remain inside the prescribed safe interval for all time. Moreover, the transformed synchronization errors converge exponentially to zero, and the original agent outputs asymptotically synchronize to a designer-selected admissible trajectory embedded in the common safe interval. Numerical simulations validate the proposed framework and demonstrate safe consensus under both asymmetric actuation bounds and time-varying output constraints. △ Less

Submitted 14 June, 2026; originally announced June 2026.

Operator Calculus for Population-Based Optimization: A Mean-Field Convergence Theory

Authors: Pekka Malo, Lauri Viitasaari, Patrik Nummi, Antti Suominen, Ankur Sinha, Olli Tahvonen

Abstract: Population-based and distributional optimization methods, from evolution strategies and consensus-based optimization to covariance-matrix adaptation and stochastic gradient methods viewed as distributional dynamics, are widely used for nonconvex or black-box problems, yet their convergence analyses remain fragmented across algorithm-specific techniques. We introduce an operator calculus in which a… ▽ More Population-based and distributional optimization methods, from evolution strategies and consensus-based optimization to covariance-matrix adaptation and stochastic gradient methods viewed as distributional dynamics, are widely used for nonconvex or black-box problems, yet their convergence analyses remain fragmented across algorithm-specific techniques. We introduce an operator calculus in which a broad class of such methods, after choosing an appropriate state space and, where necessary, augmenting the state by memory or strategy variables, is described as a composition of three elementary operators (mutation, selection, and recombination) acting on probability measures. Under explicit stability and regularity conditions, the composite operator admits a pre-generator whose continuous-time limit is a transport-reaction-jump (TRJ) PDE that preserves the operator splitting. On this foundation we establish a modular Lyapunov principle. If a state-space Lyapunov function both dissipates under the full generator and controls the relevant search-space gauges, then the state-space Lyapunov functional and the induced search errors decay exponentially. The additive generator structure allows dissipation estimates to be assembled operator by operator, providing a toolkit for certifying convergence of composite mean-field algorithms. △ Less

Submitted 12 June, 2026; originally announced June 2026.

Comments: 71 pages, 4 figures, 2 tables; ancillary files contain Python code reproducing the numerical experiments

MSC Class: 90C26 (Primary); 90C59; 35Q84; 60J25; 37N40 (Secondary) ACM Class: G.1.6

Multi-Modal Multi-Agent Robotic Cognitive Alignment enabled by Non-Invasive Consumer Brain Computer Interfaces: A Proof of Concept Exploration

Authors: Nataliya Kosmyna, Liz Jenkins, Anoop K. Sinha

Abstract: While non-verbal behaviors and expressive movements are essential for natural human-robot interaction, existing methods often overlook a crucial element: the human's internal cognitive state. Frequently, proactive multi-agent systems can interrupt humans at inopportune moments, leading to cognitive overload and decreased task performance. This paper introduces a framework for generating "cognitive… ▽ More While non-verbal behaviors and expressive movements are essential for natural human-robot interaction, existing methods often overlook a crucial element: the human's internal cognitive state. Frequently, proactive multi-agent systems can interrupt humans at inopportune moments, leading to cognitive overload and decreased task performance. This paper introduces a framework for generating "cognitively aligned" multi-agent interactions, enhancing the ability of robotic systems to contextually defer communications to the user of an agent system during moments of high human mental workload and engagement. We present the design and implementation of a closed-loop architecture that explores the interplay between autonomous task execution and real-time neurophysiological focus. Using a consumer-grade Brain-Computer Interface (BCI), our approach continuously monitors Electroencephalography (EEG) spectral band powers while a human performs an engagement-inducing task. We propose an engagement-driven pipeline where an HTTP-based signaling mechanism places a primary agent's sensory inputs and audio outputs into a holding state upon detecting high engagement. This allows secondary agents to seamlessly process complex, delegated tasks in the background. Once the human's cognitive state returns to a lower cognitive load baseline, the primary agent releases the queued agent message. Our preliminary results demonstrate the feasibility of leveraging real-time signal processing, Large Language Models (LLMs), and physical robotic embodiments to create cognitively-aware, non-intrusive multi-agent systems. △ Less

Submitted 11 June, 2026; originally announced June 2026.

Comments: 19 pages, 9 figures, for associated video, see https://youtu.be/0Tav-G87XGs

Hey Chat, Can You Teach Me? Structuring Socratic Dialogue for Human Learning in the Wild

Authors: Sidney Tio, Arunesh Sinha, Pradeep Varakantham

Abstract: Large language models are now widely used for everyday learning, but the underlying interactions are typically unstructured chats rather than following a curriculum. Unlike formal online learning systems, these interactions carry no prior record of the student, so any estimate of what the student already knows must be inferred from the dialogue itself. We show that this gap is not closed by scalin… ▽ More Large language models are now widely used for everyday learning, but the underlying interactions are typically unstructured chats rather than following a curriculum. Unlike formal online learning systems, these interactions carry no prior record of the student, so any estimate of what the student already knows must be inferred from the dialogue itself. We show that this gap is not closed by scaling models alone. Frontier and education-tuned LLMs perform poorly when asked to tutor a student over an extended session, because doing so requires three things at once. The tutor must sequence a curriculum, conduct Socratic dialogue, and infer the student's knowledge state from that dialogue. We propose separating these responsibilities. Given a student query, our system constructs a prerequisite knowledge graph in which subtopics are nodes and dependencies are edges, and frames tutoring as deciding which node to teach next and how many dialogue turns to spend on it before moving on. A lightweight PPO policy handles this sequencing decision, while an LLM conducts the Socratic exchange at the chosen node and returns a signal of student progress. Across held-out STEM and non-STEM topics, our PPO-paired tutor outperforms heuristic baselines, frontier general-purpose models, and a model specialised for Socratic dialogue: on both the rate at which students reach full curriculum mastery and the number of turns required. Explicit curriculum structure delivers gains that scaling the underlying model does not. △ Less

Submitted 10 June, 2026; originally announced June 2026.

Comments: 10 Main Body Pages, with Appendices

Mult-DPO: Multinomial Direct Preference Optimization for Recommender Systems

Authors: Yaochen Zhu, Harald Steck, James McInerney, Aditya Sinha, Yinhan He, Nathan Kallus, Jundong Li

Abstract: Direct preference optimization (DPO) is a simple and effective alignment strategy for large language models (LLMs) based on pairwise preferences. In recommender systems, however, user feedback is rarely pairwise. For a given context, e.g., a user, a session, or a conversation, we typically observe set-wise preferences with multiple positive items, where every positive item should outrank every uno… ▽ More Direct preference optimization (DPO) is a simple and effective alignment strategy for large language models (LLMs) based on pairwise preferences. In recommender systems, however, user feedback is rarely pairwise. For a given context, e.g., a user, a session, or a conversation, we typically observe set-wise preferences with multiple positive items, where every positive item should outrank every unobserved or explicitly negative item, with no prescribed order among the positives or the negatives themselves. A natural generalization is to use the Plackett-Luce (PL) reward model, which extends the Bradley-Terry reward model underlying vanilla DPO from pairwise preferences to full rankings of candidates. However, we show that adapting the PL model to set-wise preferences requires marginalizing over all positive orderings, where the resulting expression is combinatorial in complexity. To address this fundamental challenge, we propose Mult-DPO, a novel DPO objective with a tractable multinomial surrogate likelihood over set-wise preference events for the user-preference alignment of LLM-based recommender systems. The multinomial construction is not itself a ranking distribution, but it is defined on the same reward-induced weight space and admits a closed-form DPO-style objective, enabling direct alignment of LLMs with multiple candidates through a classification-style objective. In addition, we prove that the multinomial DPO loss is a tractable upper bound on the marginalized PL DPO loss when optimizing against the set-wise preference data. We further characterize the tightness of this bound in terms of the relative total weight of positives versus negatives, which provides insights into tightening the bound with richer or harder negatives. Finally, we extend Mult-DPO to the alignment of LLMs with multiple preference levels. Code is available at https://github.com/yaochenzhu/Mult_DPO △ Less

Submitted 8 June, 2026; originally announced June 2026.

Mechanistic Analysis of Alignment Algorithms in Language Models

Authors: Aarush Sinha, Ishan Garg, Veeraraju Elluru, Arth Singh, Kushal Garg

Abstract: Post-training alignment algorithms are predominantly evaluated as black boxes, obscuring how they reshape language models' internal computations. We present a systematic mechanistic analysis of six preference-optimization methods: PPO, DPO, SimPO, ORPO, GRPO, and KTO across three open-weight model families. By integrating layer-wise linear probing, Sparse Autoencoders, and crosscoders, we localize… ▽ More Post-training alignment algorithms are predominantly evaluated as black boxes, obscuring how they reshape language models' internal computations. We present a systematic mechanistic analysis of six preference-optimization methods: PPO, DPO, SimPO, ORPO, GRPO, and KTO across three open-weight model families. By integrating layer-wise linear probing, Sparse Autoencoders, and crosscoders, we localize preference representations and quantify alignment-induced geometric transformations in latent space. We find that preference signals consistently concentrate in early--mid or mid--late layers, but different objectives induce qualitatively distinct representational shifts. KTO and GRPO enhance linear separability through constructive feature sharing and sparse, high-salience recruitment. In contrast, DPO and ORPO degrade separability via non-constructive geometric rotation and feature attenuation, while PPO and SimPO largely preserve baseline geometry. These transformations exhibit architecture-dependent variability, demonstrating that behavioral alignment does not imply uniform internal restructuring. Our findings establish alignment as a heterogeneous intervention, motivate standardized feature-level auditing for safety and interpretability, and highlight the need for mechanism-aware optimization objectives. △ Less

Submitted 9 May, 2026; originally announced June 2026.

Comments: Work in Progress

Evaluation Cards: An Interpretive Layer for AI Evaluation Reporting

Authors: Avijit Ghosh, Anka Reuel, Jenny Chim, Wm. Matthew Kennedy, Srishti Yadav, Jennifer Mickel, Yanan Long, Andrew Tran, Anastassia Kornilova, Damian Stachura, Kevin Klyman, Felix Friedrich, Jeba Sania, Jan Batzner, Anoop Mishra, Eliya Habba, Yixiong Hao, Nathan Heath, Shalaleh Rismani, Usman Gohar, Andrea Loehr, David Manheim, Ruchira Dhar, Sree Harsha Nelaturu, Aarush Sinha , et al. (23 additional authors not shown)

Abstract: AI evaluation results are produced at scale but reported inconsistently across leaderboards, model cards, benchmark papers, and company blogs. The cost is interpretive: readers cannot reliably compare results across sources, identify what a report omits, or trace an aggregate claim to its underlying evidence. Recent efforts address isolated components but leave three gaps: they cover only narrow s… ▽ More AI evaluation results are produced at scale but reported inconsistently across leaderboards, model cards, benchmark papers, and company blogs. The cost is interpretive: readers cannot reliably compare results across sources, identify what a report omits, or trace an aggregate claim to its underlying evidence. Recent efforts address isolated components but leave three gaps: they cover only narrow slices of the evaluation lifecycle and do not compose into a single interpretable record; they specify static representations that do not differentiate the questions different stakeholders bring to the same evidence; and they remain proposals on paper, lacking the extraction infrastructure required for adoption at scale. We present \EvalCards{}, an operational reporting layer that composes benchmark metadata, evaluation run data, and model metadata into a unified record. We (1) derive a reporting schema from a structured review of 52 papers and 10 stakeholder interviews, (2) implement four interpretive signals (reproducibility, documentation completeness, provenance and risk, and score comparability), rendered through reader modes calibrated to research and non-research audiences, and (3) deploy a monitoring tool that applies \EvalCards{} across 5,816 models, 635 benchmarks, and 101,843 results, surfacing systematic gaps in current reporting practice. △ Less

Submitted 9 June, 2026; v1 submitted 8 June, 2026; originally announced June 2026.

How Many Counterfactuals Does It Take? Probing VLM Hallucinations Through Circuits and Causal Effects

Authors: Abhivansh Gupta, Simardeep Singh, Advika Sinha, Shreyansh Modi, Akshat Tomar

Abstract: Visual Language Models (VLMs) are known to produce hallucinated predictions that are not grounded in visual evidence, yet existing approaches lack a principled understanding of how robust such predictions are under counterfactual perturbations. In this work, we study the sample complexity of counterfactual robustness for hallucinated outputs in VLMs. We define a causal influence metric based on lo… ▽ More Visual Language Models (VLMs) are known to produce hallucinated predictions that are not grounded in visual evidence, yet existing approaches lack a principled understanding of how robust such predictions are under counterfactual perturbations. In this work, we study the sample complexity of counterfactual robustness for hallucinated outputs in VLMs. We define a causal influence metric based on log-probability differences between factual, counterfactual, and activation-patched runs, and use it to characterize the stability of hallucinated predictions. By leveraging circuit discovery techniques (CD-T), we identify model components responsible for these predictions and track their activation differences across counterfactual samples. We then derive empirical bounds on the minimum number of counterfactual samples m required to reliably detect instability in hallucinated outputs, using concentration inequalities and variance estimates of the causal influence distribution. △ Less

Submitted 7 June, 2026; originally announced June 2026.

ACM Class: I.4.3

Reshaping Undergraduate Computer Science Education in the Generative AI Era

Authors: Yi-Chieh Lee, Nattapat Boonprakong, Yugin Tan, Harold Soh, Alex Potanin, Viraj Kumar, Anoop K. Sinha, Chen Qian, Paul Denny, Mennatallah El-Assady, Ian Oakley, Jake Renzella, Amy Zhang, Jat Singh, Wee Sun Lee, Hsuan-Tien Lin, Jane L. E, Anthony Tang, Margaret M. Burnett, Sowmya Somanath, Renwen Zhang, Vicky Charisi, Alexandra I. Cristea

Abstract: Generative AI represents a turning point for Computer Science (CS) education. In recent decades, post-secondary CS education has largely focused on what has been seen as practical software engineering skills: implementation-level programming, debugging, testing, and software design, analysis, and documentation. However, this framing is becoming less tenable as generative AI automates many of these… ▽ More Generative AI represents a turning point for Computer Science (CS) education. In recent decades, post-secondary CS education has largely focused on what has been seen as practical software engineering skills: implementation-level programming, debugging, testing, and software design, analysis, and documentation. However, this framing is becoming less tenable as generative AI automates many of these tasks, challenging their centrality in CS education. To keep pace with advances in AI technology, CS curricula should consider a shift toward understanding and verifying AI-generated artifacts. This white paper outlines the findings of two international NUS-Google Workshops in Singapore, where we convened faculty members, industry practitioners, and students, and proposes a strategic response to reshape how CS should be taught at the undergraduate level. Based on the findings, we identify critical skills that must be preserved and those that are becoming less important. By incorporating these skills as "breadcrumbs," we can provide helpful nudges and engaging exercises within the current curriculum, enhancing learning experiences for everyone. We believe that to effectively prepare future computer science graduates, capable of creating, solving problems, and managing, as well as co-creating, artifacts with AI. It is important to consider a shift in curricula. Emphasizing system design, abstraction, and critical evaluation could greatly enhance their education and readiness for the challenges ahead. We propose prerequisites for solutions to reform CS education by fostering AI-native competencies, re-centering fundamental education, enhancing advanced pathways, embracing new pedagogies, and shifting institutional support. △ Less

Submitted 11 June, 2026; v1 submitted 2 May, 2026; originally announced June 2026.

Comments: Workshop report

Rapid co-design of Buoyancy-assisted robots for Challenging Locomotion using Gaussian Evolutionary Specialists

Authors: Ankit Sinha, Nitish Sontakke, Dennis Hong, Yusuke Tanaka, Sehoon Ha

Abstract: Designing high-performance legged robots requires jointly optimizing morphology and control. Model-free Reinforcement Learning (RL) offers an alternative to model-predictive control for developing robust controllers without explicitly specifying robot dynamics. Thus, we have seen theuse of RL to train controllers and evaluate designs for robot morphology optimization. While RL has shown success in… ▽ More Designing high-performance legged robots requires jointly optimizing morphology and control. Model-free Reinforcement Learning (RL) offers an alternative to model-predictive control for developing robust controllers without explicitly specifying robot dynamics. Thus, we have seen theuse of RL to train controllers and evaluate designs for robot morphology optimization. While RL has shown success inlocomotion, using it in the co-design inner loop is expensive due to repeated policy training. Universal policies conditioned on morphology offer a promising alternative, but suffer from behavioral diversity collapse, converging to a single strategy that performs sub-optimally across designs. On the other hand, end-to-end Mixture-of-Experts (MoE) architectures fail due to a collapse in its representation. We propose Gaussian Evolutionary Specialists (GES), a framework that decouples design-space partitioning from policy learning to capture diverse behaviors explicitly. GES assigns specialist policies to evolving Gaussian regions and iteratively refines them via training, probing, and territory expansion. The resulting specialists are integrated into a design sampling loop, replacing costly re-training with direct evaluation. When tested on the Buoyancy-Assisted Light Legged Unit (BALLU), GES discovers designs with 5 - 25% higher performance than naive universal policies. On hardware, a GES optimized design overcomes a 24 cm tall obstacle - 3x improvement over the baseline BALLU design. Moreover, GES curtails design optimization time by 37%. △ Less

Submitted 5 June, 2026; originally announced June 2026.

Comments: Submitted to RA-L

Terminal Time and Angle-Constrained Nonlinear Intercept Guidance

Authors: Shivam Bajpai, Abhinav Sinha

Abstract: This paper considers the problem of simultaneously controlling an interceptor's impact time and impact angle using its lateral acceleration as the sole control input. With a single control input, the nonlinear engagement kinematics is inherently underactuated, which complicates guidance law synthesis. To overcome this challenge, a hierarchical sliding mode-based guidance law is developed to concur… ▽ More This paper considers the problem of simultaneously controlling an interceptor's impact time and impact angle using its lateral acceleration as the sole control input. With a single control input, the nonlinear engagement kinematics is inherently underactuated, which complicates guidance law synthesis. To overcome this challenge, a hierarchical sliding mode-based guidance law is developed to concurrently regulate the two terminal constraints. The proposed architecture consists of a two-layer sliding manifold. The first layer comprises two sub-sliding surfaces corresponding to the impact time and impact angle error dynamics, respectively, while the second layer introduces a composite sliding manifold that combines the two individual sub-surfaces. Then, a variable-gain adaptive guidance law is designed to ensure time and angle-constrained interception against a stationary target, which is further extended to intercept a constant velocity target. Simulations are conducted for various engagement scenarios to attest to the efficacy of the proposed approach. △ Less

Submitted 1 June, 2026; originally announced June 2026.

Training Deliberative Monitors for Black-Box Scheming Detection

Authors: Aditya Sinha, Akshat Naik, Victor Gillioz, Simon Storf, Kilian Merkelbach, Rich Barton-Cooper, Axel Højmark, Marius Hobbhahn

Abstract: As autonomous agents become more capable of performing real-world tasks, distinguishing scheming behavior from benign task pursuit may become a central AI control problem. Existing monitors often rely on chain-of-thought access or internal activations, or use prompted frontier models, all of which can be unavailable, unreliable or expensive in deployment. In this work, we study action-only deliber… ▽ More As autonomous agents become more capable of performing real-world tasks, distinguishing scheming behavior from benign task pursuit may become a central AI control problem. Existing monitors often rely on chain-of-thought access or internal activations, or use prompted frontier models, all of which can be unavailable, unreliable or expensive in deployment. In this work, we study action-only deliberative monitors: smaller open-weight models trained to detect scheming and sabotage from agentic trajectories without accessing the monitored agent's reasoning or model internals. Our method, inspired by deliberative alignment, uses a scheming specification to elicit structured rationales from a frontier teacher, filters them with a separate judge, and distills the highest-quality rationales into open-weight monitors with supervised fine-tuning and reinforcement learning. We train on five datasets, and evaluate across six out-of-distribution agentic misalignment benchmarks. We show that applying our method to Qwen3.5-27B yields higher performance than all low-cost frontier models as prompted monitors (Gemini 3.1 Flash-Lite, GPT-5.4 Nano, and Claude Haiku 4.5) and than Gemini 2.5 Pro, while also achieving lower marginal inference cost (token-metered USD per 1,000 evaluations). Stronger prompted frontier monitors (Gemini 3.1 Pro, GPT-5.4, Claude Sonnet 4.6, and Claude Opus 4.6) achieve higher performance but at roughly $16$--$34\times$ higher marginal inference cost. Several of our trained monitors are positioned on the empirical cost--performance Pareto frontier among the monitors we evaluate, providing practical low-cost, low-FPR alternatives to prompted frontier models. △ Less

Submitted 28 May, 2026; originally announced May 2026.

Collaborative Threat-Aware Autonomy (CTAA)

Authors: Rajnikant Sharma, Abhinav Sinha, Isaac Weintraub

Abstract: Navigating teams of unmanned vehicles through environments containing dynamic, adversarial Weapon Engagement Zones~(WEZs) poses a fundamental challenge to mission success: a single vehicle, however capable its onboard guidance, remains a single point of failure. This paper presents a role-differentiated multi-agent framework for collaborative threat-aware trajectory planning in which a fleet of Au… ▽ More Navigating teams of unmanned vehicles through environments containing dynamic, adversarial Weapon Engagement Zones~(WEZs) poses a fundamental challenge to mission success: a single vehicle, however capable its onboard guidance, remains a single point of failure. This paper presents a role-differentiated multi-agent framework for collaborative threat-aware trajectory planning in which a fleet of Autonomous Collaborative Platforms~(ACPs) is assigned distinct roles primary intercept, escort, and decoy to improve team-level mission success probability while managing individual WEZ exposure. Each ACP independently employs a reactive guidance law derived from the Collision Sphere Boundary for Evader Zero-Set~(CSBEZ), which accounts for pursuer maneuverability constraints imposed by minimum turn radius, and steers the vehicle toward the safest heading that also makes progress toward its goal. Role assignment and spatial route separation induce two complementary effects: probabilistic redundancy, in which $N$ independent paths raise the team success probability and threat saturation, in which lower-priority escorts and decoys draw adversary attention and free the primary vehicle to transit uncontested. △ Less

Submitted 25 May, 2026; originally announced May 2026.

A Geometric Approach to Constrained Online Learning

Authors: Dhruv Sarkar, Abhishek Sinha

Abstract: We study constrained online convex optimization with adversarial time-varying constraints. At each round the learner acts before observing the loss and constraint, and is compared with the best fixed action satisfying all constraints in hindsight. The goal is to obtain minimax-optimal regret while controlling cumulative constraint violation (CCV). Prior algorithms achieved $O(\log T)$ regret with… ▽ More We study constrained online convex optimization with adversarial time-varying constraints. At each round the learner acts before observing the loss and constraint, and is compared with the best fixed action satisfying all constraints in hindsight. The goal is to obtain minimax-optimal regret while controlling cumulative constraint violation (CCV). Prior algorithms achieved $O(\log T)$ regret with $O(\sqrt{T\log T})$ CCV for strongly convex losses, and $O(\sqrt{T})$ regret with $O(\sqrt{T}\log T)$ CCV for convex losses. We present NP-OGD, an iterated nested-projection algorithm. For strongly convex losses it achieves $O(\log T)$ regret and $O(\log T)$ CCV; for convex losses it achieves $O(\sqrt{T})$ regret and $O(\sqrt{T})$ CCV. The analysis relies on a geometric movement bound: after lifting the nested projected-gradient trajectory to one higher dimension, the lifted path is self-contracted under a nonstandard norm, so a finite-length theorem for self-contracted curves controls the total projection movement. We also prove complementary lower bounds using layered sphere packings. For strongly convex losses, any online algorithm with polynomially sublinear regret can incur CCV at least $Ω((\log T)^{(d-1)/(d+1)}/\log\log T)$. For convex losses, we prove CCV lower bounds $Ω(T^{(d-1)/(2(d+3))})$ for weakly adaptive algorithms and $Ω(T^{(d-1)/(2d)})$ for NP-OGD. Finally, for the constrained experts special case over $N$ experts, an active Hedge algorithm attains $O(\sqrt{T\log N})$ regret and $O(N)$ CCV, with a matching minimax CCV lower bound for sufficiently large horizons. △ Less

Submitted 22 July, 2026; v1 submitted 20 May, 2026; originally announced May 2026.

Data-Driven Games with Coherent Risk Measures

Authors: Bharat Gangwani, Arunesh Sinha

Abstract: We introduce Coherent Utility Measure Games (CUMGs) in which players' uncertainty about the distribution of payoffs is modeled using coherent utility (risk) measures. Such measures, including mean semideviation risk and conditional value-at-risk, allow for interpretable notions of players' risk aversion while retaining formal equivalence to distributionally robust games. While CUMGs, which are a s… ▽ More We introduce Coherent Utility Measure Games (CUMGs) in which players' uncertainty about the distribution of payoffs is modeled using coherent utility (risk) measures. Such measures, including mean semideviation risk and conditional value-at-risk, allow for interpretable notions of players' risk aversion while retaining formal equivalence to distributionally robust games. While CUMGs, which are a subclass of distributionally robust games, are continuous games in general, they can be viewed as finite games ``lifted'' to the mixed strategy space, which illustrates computational challenges. Prior results extend to guarantee equilibrium existence in data-driven CUMGs. We show that the computation of approximate equilibria for CUMGs parameterized by several risk measures lies in PPAD. Consequently, we obtain finite multilinear complementarity programs for the computation of equilibrium for these games, which grow with $K$, the number of data samples. Unlike standard games, these programs are not linear in a two-player setting. Next, we establish the existence of approximate equilibria in finite data-driven CUMGs with small supports in the pure actions for the players, together with sparse data subsamples that guide the search for such equilibria. We also develop a stochastic first-order approach for smoothed CUMGs using data mini-batches, with bounds linking first-order error to approximate equilibrium. We include numerical experiments comparing the sparse-support search algorithm with complementarity-program solvers. △ Less

Submitted 16 July, 2026; v1 submitted 18 May, 2026; originally announced May 2026.

Task Capability Improvement Algorithm for Collaborative Manipulators

Authors: Keshab Patra, Arpita Sinha, Anirban Guha

Abstract: This work introduces a cooperative task capability improvement utilizing additional moments. The manipulators apply forces at the object's grasp point. Applying forces at a point other than the object's center of gravity produces undesired moments. The undesired moment acts as an additional moment. It improves the capability of an individual manipulator and, hence, the entire collaborative group.… ▽ More This work introduces a cooperative task capability improvement utilizing additional moments. The manipulators apply forces at the object's grasp point. Applying forces at a point other than the object's center of gravity produces undesired moments. The undesired moment acts as an additional moment. It improves the capability of an individual manipulator and, hence, the entire collaborative group. Any improvements in task capability directly add up to the object and transportation capability. The group's enhanced capability also helps achieve optimal capability, optimal resource allocation, and maximum fault tolerance in object manipulation. Our simulation results show an improvement in the capability of 5.86 \% compared to when no moment is used to enhance the capability of the manipulators. △ Less

Submitted 17 May, 2026; originally announced May 2026.

Improved Speed via Regional Fulfillment

Authors: Daniel Hathcock, R. Ravi, Amitabh Sinha

Abstract: In e-retail, order fulfillment speed has become one of the most important metrics affecting customer satisfaction. While common wisdom dictates that maintaining a large global fulfillment network maximizes efficiency via economies of scale, recent evidence has shown that breaking up the network into smaller regions can yield significant speed improvements. In this paper, we consider a simple abstr… ▽ More In e-retail, order fulfillment speed has become one of the most important metrics affecting customer satisfaction. While common wisdom dictates that maintaining a large global fulfillment network maximizes efficiency via economies of scale, recent evidence has shown that breaking up the network into smaller regions can yield significant speed improvements. In this paper, we consider a simple abstract model of order fulfillment by which we explain this phenomenon. We characterize fulfillment assignments satisfying an equilibrium condition based on the greedy fulfillment strategy, and quantify how the resulting fulfillment delay can be decreased by regionalizing the network. Finally, we provide some algorithmic results for computing low delay assignments, and some simulations supporting our equilibrium framework. △ Less

Submitted 13 May, 2026; originally announced May 2026.

Comments: 24 pages, 10 figures

Bounded-Input True Proportional Navigation for Impact-Time Control

Authors: Lohitvel Gopikannan, Shashi Ranjan Kumar, Abhinav Sinha

Abstract: This paper proposes a nonlinear guidance strategy capable of intercepting a constant-velocity, non-maneuvering target while strictly satisfying the prescribed bounds on the control input (commanded acceleration). Unlike conventional strategies that estimate time-to-go using linearization or small-angle approximations, the proposed strategy employs true proportional-navigation guidance (TPNG) as a… ▽ More This paper proposes a nonlinear guidance strategy capable of intercepting a constant-velocity, non-maneuvering target while strictly satisfying the prescribed bounds on the control input (commanded acceleration). Unlike conventional strategies that estimate time-to-go using linearization or small-angle approximations, the proposed strategy employs true proportional-navigation guidance (TPNG) as a baseline, which utilizes an exact time-to-go formulation and is applicable over a wide range of target motions. In contrast to most existing strategies, which do not incorporate control input bounds into the guidance design, the proposed approach explicitly accounts for these limits by modeling the interceptor acceleration as a dynamic variable. Based on the sliding mode control technique, an effective guidance law that achieves time-constrained interception while accounting for bounded input is then derived. The performance of the proposed strategy is evaluated for various engagement scenarios. △ Less

Submitted 13 May, 2026; originally announced May 2026.

Comments: Preprint; Accepted for presentation at the 15th Asian Control Conference, June 17th-21st, 2026, Indonesia

Beyond Continuity: Challenges of Context Switching in Multi-Turn Dialogue with LLMs

Authors: Aditya Sinha, Harald Steck, Vito Ostuni, Matteo Rinaldi

Abstract: Users interacting with Large Language Models (LLMs) in a multi-turn conversation routinely refine their requests or pivot to new topics. LLMs, however, often miss these topic shifts and carry over irrelevant context from previous turns, leading to inaccurate responses. In this paper, we stress-test the multi-turn understanding of LLMs and study the following two sub-tasks: (1) detecting whether th… ▽ More Users interacting with Large Language Models (LLMs) in a multi-turn conversation routinely refine their requests or pivot to new topics. LLMs, however, often miss these topic shifts and carry over irrelevant context from previous turns, leading to inaccurate responses. In this paper, we stress-test the multi-turn understanding of LLMs and study the following two sub-tasks: (1) detecting whether the user pivots or refines in the current turn, and (2) shortlisting relevant context from previous turns. To this end, we construct synthetic benchmarks based on real-world datasets from varied domains, as to simulate context shifts of different levels of difficulty. We then evaluate the zero-shot performance of ten LLMs (open-weight, closed-source and reasoning), and demonstrate that only some reasoning and strongly instructed LLMs are accurate in detecting pivots; open-weight LLMs struggle with the task and frequently carry stale context even with explicit cues; and all models suffer from a position bias. Based on the results, we discuss key takeaways for improving long-term robustness in multi-turn capabilities for LLMs. △ Less

Submitted 9 May, 2026; originally announced May 2026.

Comments: Accepted to the ICBINB Workshop @ ICLR 2026

Constrained Contextual Bandits with Adversarial Contexts

Authors: Dhruv Sarkar, Abhishek Sinha

Abstract: We study budget-constrained contextual bandits with adversarial contexts, where each action yields a random reward and incurs a random cost. We adopt the standard realizability assumption: conditioned on the observed context, rewards and costs are drawn independently from fixed distributions whose expectations belong to known function classes. We focus on the continuing setting, in which the algor… ▽ More We study budget-constrained contextual bandits with adversarial contexts, where each action yields a random reward and incurs a random cost. We adopt the standard realizability assumption: conditioned on the observed context, rewards and costs are drawn independently from fixed distributions whose expectations belong to known function classes. We focus on the continuing setting, in which the algorithm operates over the entire horizon even after the budget for cumulative cost is exhausted. In this setting, the objective is to simultaneously control regret and the violation of the budget constraint. Building on the seminal $\mathsf{SquareCB}$ framework of Foster et al. [2018], we propose a simple and modular framework that leverages online regression oracles to reduce the constrained problem to a standard unconstrained contextual bandit problem with adaptively defined surrogate reward functions. In contrast to prior works, which focus on stochastic contexts, our reduction yields improved guarantees for more general adversarial contexts, together with an efficient algorithm with a compact and transparent analysis. △ Less

Submitted 7 May, 2026; originally announced May 2026.

ViTok-v2: Scaling Native Resolution Auto-Encoders to 5 Billion Parameters

Authors: Philippe Hansen-Estruch, Jiahui Chen, Vivek Ramanujan, Orr Zohar, Yan Ping, Animesh Sinha, Markos Georgopoulos, Edgar Schoenfeld, Ji Hou, Felix Juefei-Xu, Sriram Vishwanath, Ali Thabet

Abstract: Vision Transformer (ViT) autoencoders have emerged as compelling tokenizers for images, offering improved reconstruction over convolutional tokenizers. However, existing ViT tokenizers cannot explore this landscape as performance degrades outside training resolutions, and reliance on adversarial losses prevents stable scaling. ViTok (Hansen-Estruch et al., 2025) found that the compression ratio r… ▽ More Vision Transformer (ViT) autoencoders have emerged as compelling tokenizers for images, offering improved reconstruction over convolutional tokenizers. However, existing ViT tokenizers cannot explore this landscape as performance degrades outside training resolutions, and reliance on adversarial losses prevents stable scaling. ViTok (Hansen-Estruch et al., 2025) found that the compression ratio r mediates a reconstruction-generation trade-off where lower r means better reconstructions but harder generations, so improving tokenizer reconstruction is key to more Pareto-optimal tokenizers. We introduce ViTok-v2, which addresses these limitations with native resolution support via NaFlex for generalization across resolutions and aspect ratios, and a novel DINOv3 perceptual loss that replaces both LPIPS and GAN objectives for stable training at any scale. ViTok-v2 is trained on about 2B images and scaled to 5B parameters, the largest image autoencoder to date. ViTok-v2 matches or exceeds state-of-the-art reconstruction at 256p and outperforms all baselines at 512p and above. In joint scaling experiments with flow matching generators, we show that scaling both the autoencoder and the generator advances the Pareto frontier of this trade-off. △ Less

Submitted 6 May, 2026; originally announced May 2026.

RoboKA: KAN Informed Multimodal Learning for RoboCall Surveillance System

Authors: Nitin Choudhury, Nikhil Kumar, Aditya Kumar Sinha, Abhijeet Anand, Hossein Salemi, Orchid Chetia Phukan, Hemant Purohit, Arun Balaji Buduru

Abstract: Wide exploration on robocall surveillance research is hindered due to limited access to public datasets, due to privacy concerns. In this work, we first curate Robo-SAr, a synthetic robocall dataset designed for robocall surveillance research. Robo-SAr comprises of ~200 unwanted and ~1200 legitimate synthetic robocall samples across three realistic adversarial axes: psycholinguistics-manipulated t… ▽ More Wide exploration on robocall surveillance research is hindered due to limited access to public datasets, due to privacy concerns. In this work, we first curate Robo-SAr, a synthetic robocall dataset designed for robocall surveillance research. Robo-SAr comprises of ~200 unwanted and ~1200 legitimate synthetic robocall samples across three realistic adversarial axes: psycholinguistics-manipulated transcripts, emotion-eliciting speech, and cloned voices. We further propose RoboKA, a Kolmogorov-Arnold Network (KAN)-based multimodal fusion framework designed to model structured nonlinear interactions between acoustic and linguistic cues that characterize diverse adversarial robocall strategies. RoboKA first leverages cross-modal contrastive learning to align latent modality representations and feeds the resulting embeddings to a KAN-projection head for final classification. We benchmark RoboKA against strong unimodal and multimodal baselines in both in-domain and out-of-domain setups, finding RoboKA to surpass all baselines in terms of recall and F1-score. △ Less

Submitted 30 April, 2026; originally announced May 2026.

Comments: Accepted to the International Conference on Multimedia & Expo (ICME) 2026, 7th International Workshop on Surveillance Data Processing

Agentic AI for Personalized Physiotherapy: A Multi-Agent Framework for Generative Video Training and Real-Time Pose Correction

Authors: Abhishek Dharmaratnakar, Srivaths Ranganathan, Anushree Sinha, Debanshu Das

Abstract: At-home physiotherapy compliance remains critically low due to a lack of personalized supervision and dynamic feedback. Existing digital health solutions rely on static, pre-recorded video libraries or generic 3D avatars that fail to account for a patient's specific injury limitations or home environment. In this paper, we propose a novel Multi-Agent System (MAS) architecture that leverages Genera… ▽ More At-home physiotherapy compliance remains critically low due to a lack of personalized supervision and dynamic feedback. Existing digital health solutions rely on static, pre-recorded video libraries or generic 3D avatars that fail to account for a patient's specific injury limitations or home environment. In this paper, we propose a novel Multi-Agent System (MAS) architecture that leverages Generative AI and computer vision to close the tele-rehabilitation loop. Our framework consists of four specialized micro-agents: a Clinical Extraction Agent that parses unstructured medical notes into kinematic constraints; a Video Synthesis Agent that utilizes foundational video generation models to create personalized, patient-specific exercise videos; a Vision Processing Agent for real-time pose estimation; and a Diagnostic Feedback Agent that issues corrective instructions. We present the system architecture, detail the prototype pipeline using Large Language Models and MediaPipe, and outline our clinical evaluation plan. This work demonstrates the feasibility of combining generative media with agentic autonomous decision-making to scale personalized patient care safely and effectively. △ Less

Submitted 22 April, 2026; originally announced April 2026.

Comments: 3 pages, 2 figures, submitted to ICDH IEEE conference

Privacy as Permissible Operations: An ABAC Framework for Policy-Law Compliance

Authors: Ajay Dhakar, Arunesh Sinha, Shamik Sural

Abstract: In recent years, many countries have started enacting laws to safeguard privacy of personal data of their citizens collected and maintained by various enterprises through websites, mobile apps, and other means. It is imperative that the privacy policies of these enterprises respect the provisions of the applicable law. In this paper, we show how such organizational privacy policies can be efficien… ▽ More In recent years, many countries have started enacting laws to safeguard privacy of personal data of their citizens collected and maintained by various enterprises through websites, mobile apps, and other means. It is imperative that the privacy policies of these enterprises respect the provisions of the applicable law. In this paper, we show how such organizational privacy policies can be efficiently checked against a prevalent law. Our novel approach named APLiance (\underline{A}BAC framework for \underline{P}olicy-\underline{L}aw Compl\underline{iance}) models the requirements of the different sections of a privacy law in the form of Attribute-based Access Control (ABAC) rules and the clauses of a privacy policy as a sequence of implied access requests. A policy is considered to be compliant with the law if these access requests are permitted by the corresponding ABAC rules. Although APLiance can be used in any policy-law setting, we demonstrate its effectiveness in the context of the recently introduced Digital Personal Data Protection Act of India. A browser plugin has been developed and publicly released for real time compliance checking using APLiance whenever a user visits the privacy policy page of a website. △ Less

Submitted 12 April, 2026; originally announced April 2026.

Comments: 20 pages

CONSCIENTIA: Can LLM Agents Learn to Strategize? Emergent Deception and Trust in a Multi-Agent NYC Simulation

Authors: Aarush Sinha, Arion Das, Soumyadeep Nag, Charan Karnati, Shravani Nag, Chandra Vadhan Raj, Aman Chadha, Vinija Jain, Suranjana Trivedy, Amitava Das

Abstract: As large language models (LLMs) are increasingly deployed as autonomous agents, understanding how strategic behavior emerges in multi-agent environments has become an important alignment challenge. We take a neutral empirical stance and construct a controlled environment in which strategic behavior can be directly observed and measured. We introduce a large-scale multi-agent simulation in a simpli… ▽ More As large language models (LLMs) are increasingly deployed as autonomous agents, understanding how strategic behavior emerges in multi-agent environments has become an important alignment challenge. We take a neutral empirical stance and construct a controlled environment in which strategic behavior can be directly observed and measured. We introduce a large-scale multi-agent simulation in a simplified model of New York City, where LLM-driven agents interact under opposing incentives. Blue agents aim to reach their destinations efficiently, while Red agents attempt to divert them toward billboard-heavy routes using persuasive language to maximize advertising revenue. Hidden identities make navigation socially mediated, forcing agents to decide when to trust or deceive. We study policy learning through an iterative simulation pipeline that updates agent policies across repeated interaction rounds using Kahneman-Tversky Optimization (KTO). Blue agents are optimized to reduce billboard exposure while preserving navigation efficiency, whereas Red agents adapt to exploit remaining weaknesses. Across iterations, the best Blue policy improves task success from 46.0% to 57.3%, although susceptibility remains high at 70.7%. Later policies exhibit stronger selective cooperation while preserving trajectory efficiency. However, a persistent safety-helpfulness trade-off remains: policies that better resist adversarial steering do not simultaneously maximize task completion. Overall, our results show that LLM agents can exhibit limited strategic behavior, including selective trust and deception, while remaining highly vulnerable to adversarial persuasion. △ Less

Submitted 10 April, 2026; originally announced April 2026.

Risk-seeking conservative policy iteration with agent-state based policies for Dec-POMDPs with guaranteed convergence

Authors: Amit Sinha, Matthieu Geist, Aditya Mahajan

Abstract: Optimally solving decentralized decision-making problems modeled as Dec-POMDPs is known to be NEXP-complete. These optimal solutions are policies based on the entire history of observations and actions of an agent. However, some applications may require more compact policies because of limited compute capabilities, which can be modeled by considering a limited number of memory states (or agent sta… ▽ More Optimally solving decentralized decision-making problems modeled as Dec-POMDPs is known to be NEXP-complete. These optimal solutions are policies based on the entire history of observations and actions of an agent. However, some applications may require more compact policies because of limited compute capabilities, which can be modeled by considering a limited number of memory states (or agent states). While such an agent-state based policy class may not contain the optimal solution, it is still of practical interest to find the best agent-state policy within the class. We focus on an iterated best response style algorithm which guarantees monotonic improvements and convergence to a local optimum in polynomial runtime in the Dec-POMDP model size. In order to obtain a better local optimum, we use a modified objective which incentivizes risk-seeking alongside a conservative policy iteration update. Our empirical results show that our approach performs as well as state-of-the-art approaches on several benchmark Dec-POMDPs, achieving near-optimal performance while having polynomial runtime despite the limited memory. We also show that using more agent states (a larger memory) leads to greater performance. Our approach provides a novel way of incorporating memory constraints on the agents in the Dec-POMDP problem. △ Less

Submitted 10 April, 2026; originally announced April 2026.