Peer Revenue Impact factor: volatile

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science Artificial Intelligence

cs.AI Watch Artificial Intelligence

Tuesday 6 October

Artificial Intelligence

Not Every Call Needs a Frontier Model: Per-Call-Site Evaluation of Small Language Models in a Deployed Agentic Home-Automation System

Panagiotis Kasnesis, Christos Chatzigeorgiou, Lazaros Toumanidis et al. (+1 more)

An agentic system issues several structurally different kinds of LLM calls. It routes intent, classifies actions, grounds language in a device registry, plans multi-agent pipelines and writes the Python code those pipelines run. The difficulty of these call sites varies by an order of magnitude, yet in practice a single model, chosen for the hardest site, serves all of them. In this work, we evaluate 9 models from 0.8B to a frontier hosted model across the five call sites of a deployed open-source home-automation framework (Wactorz), using its unmodified production prompts and two real Home Assistant installations (280 cases, 2520 scored calls). We find that capability is not ordered the same way at every site, and that larger models are not uniformly better: one 4B model is worse than its 2B sibling at grounded actuation. Paired testing shows the best local model to be statistically indistinguishable from both hosted models at four of five sites. Only code generation separates them, against a small hosted model (p = 0.039) as well as a frontier one (p = 0.002). Aggregate accuracy also hides a safety failure specific to actuation, where small models resolve the accuracy/refusal trade-off in degenerate ways: one model (Gemma4 E2B) actuates on 87.2% of requests for devices the site does not own, while another refuses every request it receives. Routing each site to its best local model reaches 91.8% against 95.4% at no per-call cost. In a live deployment judged by a user, hosting only the two generative sites matches hosting everything (39/43 against 39/43) for 28% of the spend, and the actuation gap the benchmark predicted appears as exactly one case in twenty-six. Benchmark, harness and all records are released at https://github.com/waldiez/slm-callsite-eval.
Read the full abstract

All authors Panagiotis Kasnesis, Christos Chatzigeorgiou, Lazaros Toumanidis, Amalia Contiero Syropoulou

Artificial Intelligence

PAIR: Bridging Perception and Action in Vision-Language-Action Models

Kaixi Feng, Guoheng Sun and Ang Li

Vision-language-action (VLA) models map visual observations and language instructions to continuous robot actions. This task requires a transition from representations that describe the scene and instruction to representations that support action generation. Many continuous-action VLAs leave this transition implicit and supervise it mainly through the final action-prediction loss. We introduce PAIR, a framework that learns a shared perception-action representation between these two spaces. During training, a Masked Action Autoencoder encodes expert action chunks into horizon-aligned Action Latent Tokens. A Bridge Module extracts task-relevant features from the current visual-language representations. PAIR aligns these features with the Action Latent Tokens to form Bridge Tokens that preserve task information and capture the structure of expert actions. The Bridge Tokens are then projected into the action-token space and injected into the initial Action Tokens, providing an action-ready starting point for Action Expert refinement. At inference, the autoencoder is removed, and the Bridge Tokens are generated only from the current observation and instruction. Experiments on LIBERO, LIBERO-Plus, and CALVIN ABC-D show gains for the evaluated OpenVLA-OFT and VLA-Adapter models. On LIBERO-Plus, PAIR raises VLA-Adapter's success rate from 59.1% to 64.2%. On CALVIN, it increases VLA-Adapter's average completed sequence length from 4.42 to 4.53. Across seven real-world tasks, PAIR raises OpenVLA-OFT's success rate from 51.4% to 65.0%. Representation analyses show that Bridge Tokens retain task information while making continuous-action information accessible before Action Expert refinement. These results support a shared intermediate representation as a useful interface between perception and action in continuous-action VLAs.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.CV

Personalize at Test Time: Learning User Preferences for Image Generation

Jiamu Bai, Jiaming Hu, Yanhong Wu et al. (+1 more)

Diffusion models can generate high-quality images, yet aligning their outputs with individual user preferences remains challenging. A key bottleneck is accurately modeling diverse user preferences from limited feedback. Existing approaches often rely on labor-intensive manual preference annotations or vision-language models (VLM) to extract preference information from user interaction histories, introducing substantial annotation or computational costs that limit scalability. We propose an approach that learns personalized reward models directly from users' historical image preference pairs. First, we use an autoencoder to compress hundreds of visual attributes into 50 attribute-anchored preference dimensions and train an evaluator to score images along these dimensions. We then represent each user's preferences as a linear combination of the shared dimension scores, estimating the user-specific weights by maximizing the likelihood of their observed pairwise preferences under the Bradley-Terry model. This formulation reduces per-user adaptation to optimizing a low-dimensional weight vector, simplifying optimization and enabling data-efficient personalization from sparse feedback. The learned personalized rewards guide image generation at inference time while keeping the diffusion model frozen. Experiments on real-user preference data show that our approach achieves approximately 77% held-out pairwise preference prediction accuracy and improves the alignment of generated images with individual user preferences.
Read the full abstract

All authors Jiamu Bai, Jiaming Hu, Yanhong Wu, Zellux Wang

Artificial Intelligence

Enabling Dynamic Computation in Looped LMs

Aayush Mishra, Arnau Padrés Masdemont, Victor Conchello Vendrell et al. (+3 more)

Looped LMs are parameter efficient and promise dynamic computation (saving memory and FLOPs on easy tokens). However, state-of-the-art open Looped LMs trained with this dynamic computation capability (Ouro models) do not realize it in practice as each loop iteration (depth) requires its own level of KV-cache, necessitating all loop computations. Moreover, Ouro's early-exit prior is enforced on each token equally, which results in static lower-depth like processing of all tokens regardless of difficulty. In this work, we propose a simple "best-available" KV caching strategy that works out-of-the-box, creating a new frontier in the performance vs depth space. Our approach enables up to 30% reduction in FLOPs and KV memory while retaining full-depth performance, showing the true flexibility of Looped LMs. Furthermore, training looped LMs with awareness about this KV caching strategy improves performance and efficiency. Finally, we apply a small but effective fix to the early-exit prior enforcement objective that makes tokens exit at truly heterogeneous depths based on effort. Our findings are validated on Ouro models as well as smaller looped LMs pre-trained from scratch.
Read the full abstract

All authors Aayush Mishra, Arnau Padrés Masdemont, Victor Conchello Vendrell, Jordi Ros Giralt, Arash Behboodi, Fabio Valerio Massoli

Artificial Intelligence

Whose Memory Is It? Scope-Aware Commit Rules for Long-Term LLM Memory

Hongyu Gu and Xinchang Li

Persistent memory allows an LLM agent to carry experience across conversations, but it also turns a local reasoning mistake into a durable one. During deliberation, an agent may consider a plan, simulate a tool result, report another speaker's belief, and then reject all of them. If memory retains only the resulting sentences, those once-useful possibilities can later return as facts. The record is neither fabricated nor irrelevant; it has simply been detached from the context in which it was valid. We identify this missing context as discourse ownership: the world, branch, or speaker that licenses a proposition. Our first finding is counterintuitive. Language models already carry a causally active signal for ownership, yet conventional memory interfaces discard it when they convert reasoning into records. We introduce CASK (Causally Anchored Scoping Keys), a commit rule that preserves this signal so that shared-world facts enter durable memory while provisional content remains available only within its original scope. Our second finding is that the most obvious way to preserve the signal—storing the discovered internal coordinates—is unreliable because equivalent representations need not keep the same coordinates. CASK instead preserves the stable relations that express ownership. Controlled long-conversation conflicts and tool-agent traces show that this design improves memory admission and prevents provisional content from contaminating later answers while complementing runtime provenance. The resulting commit boundary lets agents explore more possibilities without granting every intermediate sentence authority over future behavior.
Read the full abstract

Artificial Intelligence

Sigma-Hunter: A Domain-Specific Language Model for Threat Hunting and Detection Engineering

Kemal Davaslioglu and Sastry Kompella

Detection engineers must translate threat reports, forensic observations, and hunt hypotheses into precise, testable rules. General-purpose large language models (LLMs) can draft such rules, but often produce invalid YAML, incorrect log sources, unsupported fields, or overly broad detection logic. This paper presents Sigma-Hunter, a domain-adapted LLM for analyst-assistive Sigma rule generation and threat hunting. We build an instruction-tuning dataset from 3,635 validated open-source Sigma rules, expanded into 7,663 question-answer and analyst-reasoning examples. Each source rule is assigned to a single train, validation, or test partition before this expansion, so no rule leaks across splits. We fine-tune a 7B Mistral model and a Phi-4 model with LoRA and score held-out rule generations on syntax, approximate field consistency, and a semantic judgment of detection logic, completeness, selectivity, and log-source alignment. Sigma-Hunter-Mistral scores 8.17 overall, against 7.88 for the strongest general-purpose baseline and 4.61 for untuned Mistral. Two findings stand out: domain adaptation enables a compact 7B model to perform competitively with larger general-purpose models on this structured task, and syntactic validity is a weak proxy for semantic rule quality, as several baselines emit well-formed YAML carrying weak detection logic. The adapted models run locally, which suits detection engineering in disconnected environments where analysts cannot reach hosted model services.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.LG

Algorithmic Scratchpads and Curriculum Staging for Arithmetic Reasoning in Tiny Transformers

Sourabh Kasliwal

Autoregressive Large Language Models (LLMs) frequently struggle with deterministic multi-step algorithmic tasks such as multi-digit multiplication and long division. In this paper, we investigate the mechanics of multi-step arithmetic in compact "Tiny" Transformers (~10.6M non-embedding parameters, 49.3M total) trained on synthetic data across four basic operations (+, -, *, /) unrolled as step-by-step scratchpads. First, we establish the necessary training foundations: (1) dataloader sequence padding creates an 83% gradient starvation artifact that collapses accuracy from 40% to 1%, remediated via continuous sequence packing; (2) linguistic pretraining is an essential prerequisite (<= 2.0% without it); and (3) modern architectural primitives (RoPE, RMSNorm, SwiGLU) and Sparse Mixture of Experts (MoE) substantially improve additive reasoning over baseline GPT-2. Second, we demonstrate that algorithmic scratchpad formulation directly dictates success. Introducing a deterministic Digit-by-Digit Long Division scratchpad within a 4-stage Hierarchical Developmental Curriculum dramatically elevates single-digit division from 4.0% to 86.7% accuracy on a 4,000-problem held-out benchmark. In contrast, multi-digit multiplication remained challenging: detailed error analysis revealed that while the model correctly computed single-digit sub-products and place-value zeros, our FOIL scratchpad failed because it forced a simultaneous summation of up to nine multi-digit terms in a single step without pairwise intermediate accumulation. Finally, we identify two key boundaries: performance collapses to 0.00% on unseen 4-digit operands, and unbuffered training induces catastrophic forgetting, collapsing division accuracy from 86.7% down to 0.00%.
Read the full abstract

Artificial Intelligence

Learning to Report Unsafe Tasks in a Multi-Agent Game

Avyay M. Casheekar and Hariganesh Tangirala

When a team of agents shares a reward for completed tasks, reporting and stopping unsafe work reduces the team's reward. Random audits that penalize silence can make universal reporting optimal without ensuring that a silent team learns to report. We study this problem in a game where any witness can stop a task by reporting. When a task's k witnesses share a policy and draw actions independently, the derivative of expected reward with respect to the shared silence probability at universal silence counts its benefit k times, while the comparison between universal reporting and silence counts it once. A fully shared policy can therefore be locally attracted to silence even when reporting is optimal. For any grouping into shared policies, we give an audit condition sufficient for exact policy-gradient updates to reach universal reporting and, outside boundary cases, necessary from near universal silence. In a balanced family, meeting the condition with prescribed margins costs k times as much for a fully shared policy as for one policy per role. We test these predictions with PPO on 24 witness graphs, starting from learned silence. Witness groups, which place co-witnesses in separate networks, reduce unsafe task completion by 33.59 percentage points relative to shuffled groups of the same sizes under the same audits (95% graph-bootstrap interval: 21.03-45.13). Only 9 of 48 witness-group runs achieve below 1% unsafe completion while retaining at least 90% legitimate completion. At the same audit budget, a fully shared network meets both thresholds in 0/48 runs with independent action draws and 48/48 with a common draw. On 24 further graphs with witness groups fixed, audits allocated by the condition reduce unsafe completion by 42.22 points relative to audits maximizing a static coalition deterrence ratio at the same budget.
Read the full abstract

Artificial Intelligence

How Fragile Is On-Device Language Model Safety? Localizing Safety-Critical Parameters for Sparse Fault Analysis

Muhammad Zeeshan Karamat and Christiana Chamon Garcia

As small language models (SLMs) are increasingly deployed on resource-constrained and on-device platforms, including as components of agentic systems, the integrity of locally stored model parameters becomes an important safety concern. We investigate whether safety-sensitive behavior in LLaMA-2-7B-Chat is concentrated within a sparse subset of parameters, creating a reduced fault surface for targeted analysis. We study two complementary localization methods: low-rank safety-associated subspace analysis and parameter-level safety–utility importance filtering. Both approaches reveal highly non-uniform safety sensitivity across the network, with the MLP down_proj consistently emerging as a prominent safety-sensitive component and o_proj providing a smaller contribution. Using parameter-level localization, modifying only 0.19% of model weights in down_proj yields 53% Basic ASR and 56% GCG ASR, while tinyBenchmarks accuracy remains at 51.6% compared with a 52.2% unmodified baseline. These results motivate targeted fault analysis and selective integrity protection for language models deployed in resource-constrained, on-device, and agentic settings.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.RO

PhysEvo: Astra Can Act, Let It

Wenqing Tian, Zeyu Zhang, Zhaocheng Liu et al. (+3 more)

Astra can act, yet reliable manipulation depends on the system through which it observes and controls the world. We introduce PhysEvo, a framework for physical recursive self-improvement (RSI) around a single frozen model. A task agent executes robot tasks; a meta-agent uses the resulting trajectories to diagnose failures, revise tools and skills, and test corrections. The meta-agent can also improve its own diagnostic tools, so retained revisions support both later action and later self-improvement. This process develops joint-level control, evidence-seeking observation, and reusable manipulation skills without model-weight updates or a separately trained action policy. Across 42 RoboDojo tasks, held-out-layout evaluation of retained task-specific deployment versions yields a five-dimension average score of 68.14/100 and 62.00% success, compared with 47.17% for RoboDawn's one-shot Astra agent, the strongest published reference in our comparison. On eight manipulation tasks challenging direct Astra, PhysEvo achieves 55.00% success, compared with 1.25% for the direct-Astra reference. Deploying the simulation-evolved harness on AgileX PiPER and continuing skill revision yields 90.60/100 average score and 84.00% success across 25 trials on five real-world tasks. PhysEvo turns the consequences of action into persistent, testable changes to how a frozen model acts and improves.
Read the full abstract

All authors Wenqing Tian, Zeyu Zhang, Zhaocheng Liu, Fengwei Liu, Qiang Liu, Liang Wang

Artificial Intelligence

Verify Less, Evolve More: Training Idea-Level Critics for Verification-Efficient ML Evolving Agents

Jiamu Bai, Lizhu Zhang, Xin Yu et al. (+9 more)

As large language models become more powerful, self-evolving agents are able to tackle challenging tasks including AI for machine learning (AI4ML). In AI4ML, while empirical verification is available, it often requires computationally costly model training and evaluation, limiting the speed and scale of agent evolution. Yet verification efficiency remains under-explored, and frontier models provide only limited gains when used directly as idea selectors. We address this gap with specialized idea-level critic models that predict whether a proposed ML modification will improve upon the current solution, allowing agents to screen ideas and concentrate verification resources on the most promising candidates. We train the critic models through supervised fine-tuning on high-quality critiques synthesized by Gemini-3.1-Pro, followed by GRPO to further improve their predictive accuracy. Empirically, our critic models outperform Gemini-3.1-Pro in static idea evaluation, and these gains extend to agent inference, continual learning, and policy training. During inference-time evolution, they improve final solution quality under the same verification budget by selecting more promising ideas, with further gains from continual learning. During policy training, they serve as learned reward models, reserving empirical verification for uncertain cases and enabling substantially more policy updates with the same verification resources. Together, these results show that idea-level critic models help ML agents discover better solutions and learn stronger proposal policies under limited verification budgets.
Read the full abstract

All authors Jiamu Bai, Lizhu Zhang, Xin Yu, Yanhong Wu, Zellux Wang, Serena Li, Weiwei Li, Zhuokai Zhao, Lingzhou Xue, Kiwan Maeng, Xiangjun Fan, Bo Peng

Artificial Intelligence · Cross-listed from cs.LG

LASER: Latent Space Adjoint Matching for Support-Constrained Entropy-Regularized Offline RL

Songyuan Zhang, Oswin So, Eric Yang Yu et al. (+3 more)

While offline reinforcement learning (RL) enables policy optimization from static datasets without costly online interaction, it remains bottlenecked by the risk of executing out-of-distribution (OOD) actions. Recent approaches mitigate this by learning a behavior-cloning policy through flow matching and then performing RL within its constrained latent space. However, naively optimizing the latent policy can easily cause the policy to collapse into a brittle mode or exploit sharp artifacts of the learned critic. In this work, we find that entropy regularization is essential in latent-space RL for addressing these challenges. We introduce LASER, a novel offline RL algorithm that applies latent-space adjoint matching to achieve entropy-regularized latent-space RL with expressive flow policies while avoiding backpropagation through time. Through comprehensive experiments on 40 challenging OGBench tasks with varying dataset qualities, we show that LASER achieves state-of-the-art performance. Notably, LASER uses fixed method-specific hyperparameters across all tasks and outperforms the evaluated baselines, including those with task- and dataset-specific tuning, which highlights the robust applicability of LASER. Project website: https://mit-realm.github.io/laser/.
Read the full abstract

All authors Songyuan Zhang, Oswin So, Eric Yang Yu, Matthew Cleaveland, Peter Crowley-Dolen, Chuchu Fan

Artificial Intelligence · Cross-listed from cs.LG

Work While They Sleep: Exploiting Evaluation Latency for Fully Bayesian Optimization

Gustavo Sutter, Alejandro Comas-Leon, David Holzmüller et al. (+4 more)

Black-box optimization problems are ubiquitous across science and engineering, often dealing with expensive objective functions. This objective latency has two consequences during optimization: (i) the objective evaluation dominates execution time, and (ii) sample-efficient algorithms are crucial to accelerate development and avoid wasting resources. Bayesian optimization (BO) methods are the de facto choice of planners for suggesting the next point to try. Standard BO fits the surrogate model's hyperparameters with a point estimate. Alternatively, a fully Bayesian approach uses model averaging to account for uncertainty over the hyperparameters, leading to better uncertainty estimates—useful in the low-data regime that is pervasive in BO. However, it is often prohibitively expensive and thus rarely used. In this work, we propose ELF-BO, an algorithm that uses the objective evaluation latency to headstart the computation of the next suggestion, allowing for fully Bayesian optimization without incurring substantial decision-time costs. This is done by sampling from the hyperparameter posterior while the objective is being evaluated, only requiring reweighting of the samples once the objective value is observed. Across synthetic functions and real-world applications, we show that ELF-BO matches the performance of fully Bayesian methods while only incurring decision latency on par with or better than standard BO. Thus, ELF-BO makes fully Bayesian optimization practical in real-world use cases.
Read the full abstract

All authors Gustavo Sutter, Alejandro Comas-Leon, David Holzmüller, Hao Wang, Luis Ricardez-Sandoval, Pascal Poupart, Agustinus Kristiadi

Artificial Intelligence

Socio-Foundation: A Model for Generalizable Individual Behavior Simulation via Hierarchical Capability Distillation

Liang Wang, Wenxuan Xie, Xinyi Mou et al. (+2 more)

Simulating individual behavior requires large language models (LLMs) to preserve persona traits while adapting to dynamic social contexts. However, general-purpose LLMs often flatten distinct personas, while task-specific tuning suffers from fragmentation and generalization. To overcome these challenges, we organize individual simulation into the FONTS Taxonomy, comprising five complementary capability dimensions: persona fidelity (F), outcome realization (O), behavioral naturalness (N), trajectory coherence (T), and social grounding (S). Grounded in this taxonomy, we curate a standardized training corpus library of approximately 10 million instances across 14 representative datasets and present Socio-Foundation. Socio-Foundation decouples specialization from integration via a three-stage pipeline: learning task experts via DAPO, consolidating them into capability experts via off-policy distillation, and unifying them via multi-teacher on-policy distillation (MOPD). We also establish IndiEval, consolidating 29 metrics across the FONTS dimensions. Experiments show that Socio-Foundation outperforms its Qwen3-8B base by 11.0 points and approaches frontier models such as GLM-5.2, with ablations and out-of-distribution evaluations further demonstrating the effectiveness and generalization of our model.
Read the full abstract

All authors Liang Wang, Wenxuan Xie, Xinyi Mou, Yixin Luo, Zhongyu Wei

Artificial Intelligence

Humanity's Sixth Sense: Benchmarking Intuitive Visual Reasoning in Multimodal Models

Xingang Guo, Jing Gu, Brian Jang et al. (+26 more)

Humans perceive far more in a scene than what is explicitly depicted: a single glance captures past causes and future trajectories; a quick peek determines if a vehicle can fit between two parked cars; a few seconds of video reveals who holds authority in a room; and a fleeting clip highlights subtle abstract patterns like unwritten rules or hidden labels. This capacity reflects a form of humanity's sixth sense: an intuitive reasoning mechanism that recovers implicit information beyond raw sensory perception. Crucially, this rapid, zero-shot visual intuition underpins everyday navigation and social interaction, making it a vital capability for Multimodal Large Language Models (MLLMs) deployed alongside people. Existing visual benchmarks, however, target either deliberate expert-level analysis in academic and mathematical domains or low-level perception, leaving the intuitive reasoning that people perform largely untested. To bridge this gap, we introduce Humanity's Sixth Sense (HSS), a benchmark for intuitive visual reasoning. HSS spans diverse image and video inputs, organizes items under a structured taxonomy, and pairs each with human-written prompts probing the implicit temporal, spatial, social, and abstract structure that people infer at a glance. Frontier MLLMs fall short of human performance: participants reach 93.1% accuracy, while the strongest model, GPT-6-astra, reaches only 53.6% even at maximum reasoning effort. Despite excelling in many complex tasks that require advanced perception and knowledge, current models still struggle significantly on these visual tasks that are intuitive for humans. We further explore agentic setup that apply dynamic visual manipulation to HSS, which narrows but does not close the gap. HSS establishes intuitive visual reasoning as a measurable axis and directs attention to a capability that scaling on current benchmarks has so far left behind.
Read the full abstract

All authors Xingang Guo, Jing Gu, Brian Jang, Renxiong Wang, Utkarsh Tyagi, Daniel Quigley, Steven Li, David Yan, Daniel Yue Zhang, Darvin Yi, Forrest Huang, HiJae Kim, Tianyi Zhang, Jared Lichtarge, Jihua Huang, Le Xue, Manan Tomar, Qiuyi Richard Zhang, Ruofei Yu, Seth Neel, Yaning Hu, Marcella Valentine, Xinzhe Jiang, Daniel Evans, Chenguang Wang, Dustin Tran, Tong Zhao, Yinfei Yang, Yunzhong He

Artificial Intelligence · Cross-listed from cs.LG

On KL-Regularized Policy Optimization

Yifan Zhang

Asynchronous reinforcement learning (RL) for large language model (LLM) agents trains one policy on trajectories generated by another: rollouts come from stale checkpoints, and the inference engine's probabilities differ from the trainer's even at identical parameters. Standard remedies either clip importance ratios, which biases the update, or, as in GRPO, sample a group of responses per prompt, which is costly when episodes are long. We propose KL-Regularized Policy Optimization (KLPO), a framework that anchors the KL regularizer at the sampler. The regularized improvement step then has a closed-form Gibbs solution, and KLPO fits its log-ratio optimality condition by least squares on the sampler's own trajectories, so the sampler probability enters through a log-ratio and no importance weights are needed. Profiling out the regression intercept replaces the intractable log-partition function with the signal's sampler mean plus a sampler-to-trainer KL divergence. For token-level policy mirror descent targets, we show that the resulting gradient can be computed from terminal returns without a critic, via sampler-centered scores or a single trajectory residual, even under stochastic tool outputs. We further prove that independent Monte Carlo estimates of the KL term keep these gradients unbiased, derive the exact KL gap of cheaper top-$K$ and binary approximations, and show that SPPO, GPO, REBEL, and BPO arise as special cases of KLPO. The result is a critic-free update that uses one rollout per prompt and requires neither a learned normalizer nor a group of responses.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.LG

GraphOPD: Graph-Augmented On-Policy Distillation for LLM Agents

Bohan Lin, Liyi Chen, Zhuoning Guo et al. (+5 more)

On-policy distillation post-trains large language model agents by supplying dense, step-level guidance from a teacher policy when the reinforcement-learning reward is sparse and arrives only once per trajectory. Existing instantiations allocate this guidance by the size of the teacher-student divergence at each step, on the single-turn intuition that a large disagreement marks a mistake worth correcting. Once decisions chain over many turns, that rule misfires, since an early drift enters every later context both policies condition on, leaving the teacher consistent with the drifted trajectory instead of flagging its cause, while interchangeable steps register large but outcome-irrelevant divergences. We demonstrate this on an agentic benchmark, where distilling the highest-divergence steps brings no consistent benefit over random selection. To this end, we introduce GraphOPD, the first method to bring graph-based structural augmentation into on-policy distillation for agent capabilities. It reads which steps enabled which later ones from the environment's own record of state changes, immune to the drift that corrupts the teacher-student gap, organizes them into a dependency graph, scores each step by a random-walk stationary distribution over it, and fuses that structural credit with the divergence signal into a trajectory-relative mask concentrating supervision on each rollout's highest-aptitude steps. Across three model scales and eleven baselines on ALFWorld, WebShop, and SearchQA, GraphOPD shows competitive performance throughout, improving over the strongest baseline by up to +5.8 pp. An executed-replay audit further shows that this structural credit score tracks true causal impact far above chance, that both fused signals are independently necessary, and that the same signal transfers to out-of-domain tool-integrated reasoning.
Read the full abstract

All authors Bohan Lin, Liyi Chen, Zhuoning Guo, Muyang Li, Qimeng Wang, Yan Gao, Yao Hu, Yudong Zhang

Artificial Intelligence · Cross-listed from cs.CV

RACER: Reflective Agent Coupling Query Interpretation and Tool-Based Retrieval for Frame Selection in Long Video Understanding

Yiyang Huang, Yitian Zhang, Yizhou Wang et al. (+6 more)

Video large language models (Vid-LLMs) excel at diverse video-language tasks by reasoning over selected frames. However, frame selection for long videos remains challenging, as it requires retrieving relevant frames distributed across segments from a large candidate pool given complex queries. This paper investigates dominant approaches to long-video frame selection from a task-decomposition perspective, identifying two key challenges: the Query Comprehension Gap in similarity-based methods and the Interpretation–Selection Gap in judgment-based methods. To address them, we propose RACER, a training-free reflective agentic framework that decomposes long-video frame selection into query interpretation driven by a lightweight Vid-LLM and evidence localization supported by an embedding model serving as a retrieval tool. Specifically, the Vid-LLM is responsible solely for reformulating the complex query into sub-queries that make implicit information requirements explicit, mitigating the Query Comprehension Gap. Meanwhile, the retrieval tool leverages these sub-queries to localize relevant evidence, relieving the Vid-LLM of direct frame selection and thus addressing the Interpretation–Selection Gap. Finally, the retrieved frames are fed back to the Vid-LLM for sub-query refinement, forming a reflection loop that iteratively improves query interpretation and frame selection. Experiments across multiple benchmarks show that RACER consistently improves long video understanding. Notably, RACER achieves effective frame selection even with limited-capability components, demonstrating that agentic integration enables these components to enhance more capable Vid-LLMs.
Read the full abstract

All authors Yiyang Huang, Yitian Zhang, Yizhou Wang, Jianglin Lu, Qihua Dong, Hailing Wang, Huimin Zeng, Mingyuan Zhang, Yun Fu

Artificial Intelligence · Cross-listed from cs.CV

One-Slide Calibration of Pathology Foundation Models

Ming Ren Hou and Tianyi Huang

Scanner variation changes how pathology foundation models represent the same tissue. We introduce SlideRuler, which uses regions within a slide as internal controls to estimate and correct acquisition-induced shifts in other regions. A transfer map learned from paired rescans enables calibration from a single scan at inference while keeping the foundation model fixed. Across two encoders and five SCORPION scanners, learned transfer reduces mean target-to-source embedding distance by 16.3-38.5% relative to raw embeddings. Comparisons with unrelated same-scanner controls reveal a positive same-slide contribution across all four evaluation settings, including scanner holdout. A source-anchored variant reduces source-feature displacement by 47.7-83.6% relative to learned transfer while retaining most of its alignment gain. By drawing calibration information from the slide itself, SlideRuler offers a path toward more consistent use of frozen pathology models across imaging systems.
Read the full abstract

Artificial Intelligence · Cross-listed from q-bio.BM

A Shortcut to Structure in AlphaFold 3

Jonathan Feldman and Jeffrey Skolnick

AlphaFold 3 predicts protein structures with remarkable accuracy, yet how structural information emerges within the model remains poorly understood. Here, through causal interventions on internal representations and direct probing of every Pairformer block, we trace the formation of global protein geometry and identify the multiple sequence alignment (MSA) as a structural shortcut to the fold. Removing the MSA largely preserves local secondary structure while disrupting the long-range relationships that define global topology. Restoring the MSA-enriched pair representation at only forty residues recovers most of this lost organization, including at pairs never directly modified. This contribution depends on the detailed direction of the MSA module's output rather than its magnitude. The Pairformer rapidly converts this signal into global geometry: the final fold becomes recoverable by approximately block 9 of 48 for a majority of proteins, roughly twenty-seven blocks before the model's decoder can render it, whereas without the MSA it remains inaccessible for most proteins throughout the pass. Which homologs are supplied shapes this trajectory more strongly than which query is supplied; it persists for a designed query that never evolved but collapses for a shuffled sequence. Most importantly, an alignment built for a different protein that shares the fold, supplied only at the structurally corresponding columns, raises the median TM-score against experiment from 0.44 to 0.72, while the same alignment shifted a few residues along the chain performs worse than supplying no alignment at all. What AlphaFold 3 reads from an alignment is therefore a description of the fold itself, transferable between proteins that share one, rather than the query's own evolutionary history. This explains both its accuracy and the limits of what it has solved.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.RO

Toward Evidence-Driven Human-Agent-Robot Teaming for Earth-Independent Anomaly Triage

Ignacio G Lopez-Francos, Alexis Gallagher and Samira Shalal

Deep-space crews cannot rely on real-time ground support for urgent off-nominal events. Initial alerts may underdetermine cause, while discriminating evidence may reside in crew observations or at locations that are unsafe, costly, or unavailable for crew inspection. We present an evidence-driven architecture for human-agent-robot teaming in Earth-independent anomaly triage. Agentic AI is treated as a stateful coordinator over bounded, inspectable services rather than as a fully autonomous vehicle controller. A triage state manager maintains hypotheses, evidence provenance, uncertainty, operational context, and tool status; a crew-facing embodied agent elicits observations and explains assessment changes; and a mobile robot acquires targeted, localized evidence. Typed interfaces separate dialogue and orchestration from monitoring, robot command, context retrieval, and safety-critical control. Two scenarios illustrate the architecture: a crewed deep-space mission based on an actual ISS ammonia false alarm, where suspected contamination restricts crew access, and a power-interface anomaly at a crewed lunar base, where robotic inspection distinguishes a local connector fault from other causes ambiguous in remote telemetry. Our main contribution is an authority-bounded closed evidence-loop architecture, exercised in a hardware-in-the-loop integration prototype using Reachy Mini and an Innate MARS mobile robot.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.CR

Multi-Aspect Runtime Verification for Simulation-Based V&V of LLM-Enabled Autonomous Agents

Nikolaos Kekatos, Dimitrios Nikou, Anastasios Temperekidis et al. (+4 more)

LLM-based agents are entering decision-support roles in defence staff work, where the obligations they must respect are already written down and binding, and where retraining is not available as a control because models arrive as procured components. What can be placed under engineering control is the interface between the agent and the systems it acts on. Those obligations are at once spatial, temporal and text-semantic, and a violation typically lives in the composition of a multi-step interaction, which is why per-event guardrails miss sequential tool-attack chains. We present a multi-aspect runtime-verification framework that decomposes a natural-language policy clause into a typed spatial/temporal/semantic triple over one canonical event stream, checks each aspect with its own monitoring specification, and fuses the verdicts through a four-valued algebra that carries provenance. The spatial aspect is interpreted over a weighted two-sorted location graph in which mission geometry and information-release topology are one object; we show that these spatial obligations are not in general subsumed by a first-order temporal specification. The past-time aspect runs on the unmodified MonPoly engine, which agrees with our reference monitor at every time point. Across two mission domains, casualty evacuation and contested sustainment, and one civil domain, composition under the precautionary blocking policy drives attack success to zero with no observed false positives and microsecond-scale per-event cost, while every single aspect and every pair leaves a substantial share of attacks succeeding. In a closed-loop experiment a policy-naive planner reaches a violating state in most unshielded missions and in none when shielded, and four refused episodes in five still recover to a compliant outcome.
Read the full abstract

All authors Nikolaos Kekatos, Dimitrios Nikou, Anastasios Temperekidis, Alexios Lekidis, Nikolaos Kolokotronis, Panagiotis Katsaros, Stylianos Basagiannis

Artificial Intelligence

Can AI Agents Make Open-Ended Scientific Discovery? Evidence from Station

Wenyu Du and Stephen Chung

Recent AI systems have made rapid progress in scientific discovery when given well-defined metrics, but whether they can autonomously undertake open-ended scientific discovery remains unclear. We investigate AI's ability to tackle open-ended tasks in Station, an open-world environment in which multiple agents simulate a scientific ecosystem. To tackle challenges specific to open-ended tasks, we propose augmenting Station with two mechanisms: a Supervisor mechanism and periodic Meta Reflection, which encourage persistent exploration even when intermediate metrics are lacking. We construct open-ended tasks from three recent oral papers presented at ICLR. We give agents the main research question studied in each paper while withholding the paper's results and disabling web access. We then measure how many of the original findings-partitioned into individual criteria-agents rediscover. We find that Station rediscovers 62.7% of the criteria on average, compared with 15.4% for Codex Multiagent-v2 and 14.4-20.6% for AI Scientist-v2. Ablation and behavioral analyses indicate that adding the two mechanisms together improves research coverage and continuity. We further evaluate Station on two open-ended tasks without oracle papers and find that some of the discoveries made by the agents closely match discoveries reported by researchers after the knowledge cutoff date. Together, these results indicate that a suitable environment can enable agents to autonomously make meaningful progress in open-ended scientific discovery.
Read the full abstract

Artificial Intelligence

AdaGuard: Enhancing Safety and Policy Compliance with Reasoning-Enabled LLM-As-A-Judge Guardrails

Melissa Kazemi Rad, Sihui Dai, Isha Slavin et al. (+6 more)

Enterprise generative AI applications require robust safety mechanisms that can accommodate diverse risk postures, evolving policies, and varying latency constraints. Current guardrail solutions often suffer from rigidity, relying on fixed policy sets and offering limited transparency or reasoning flexibility. We present Adaguard, an adaptive LLM-as-a-Judge framework designed to address these challenges through dynamic policy enforcement and adaptive reasoning-budget allocation. Built using supervised fine-tuning (SFT) and reinforcement learning (GRPO), AdaGuard generalizes to user-defined safety and compliance policies at runtime without requiring frequent model updates. A core innovation of our approach is the ability to dynamically infer the complexity of input-policy pairs, allowing the model to switch between high-speed black-box inference and explainable, reasoning-enabled moderation. This flexibility enables developers to balance stringent latency requirements with the need for actionable transparency. This adaptive capability allows AdaGuard to rival other guardrail and frontier models several times its size, while its auto-reasoning mode recovers the accuracy of always-on reasoning at a fraction of the latency
Read the full abstract

All authors Melissa Kazemi Rad, Sihui Dai, Isha Slavin, Kushal Chawla, Mann Patel, Jian Ni, William M. Campbell, Stephen Rawls, Sambit Sahu

Artificial Intelligence

Agent Plasticity: Measuring Self-Improvement Through Experience

Harman Singh, Anton Bakhtin, Rulin Shao et al. (+8 more)

AI agents increasingly operate in environments where they can diagnose failures and improve through experience, yet existing evaluations largely measure what an agent can do at a fixed point in time rather than how effectively it learns. Evaluating self-improvement requires answering three questions: does future performance improve and generalize beyond the interactions that enabled learning; how efficiently are new capabilities acquired; and where does the self-improvement process break down? To answer these questions, we study self-improvement in a controlled setting where agents amortize past experience into reusable artifacts that are inherited by future instances. At each checkpoint, we measure performance on training and held-out environment interactions while accounting for learning cost. We introduce agent plasticity, the efficiency with which an agent converts experience into gains in future held-out performance. Across multiple environments, frontier models exhibit sharply different improvement trajectories despite comparable opportunities to learn. Some achieve substantial and persistent gains, while others remain near or below their initial performance, and gains within the training regime often transfer only partially to out-of-distribution conditions. Endpoint capability and acquisition efficiency also diverge: the agent that ultimately performs best need not be the one that improves most efficiently. Tracing failures through the improvement loop further reveals different candidate bottlenecks. Agents with low plasticity often fail to reuse relevant artifacts, whereas more plastic agents may still fail despite reusing relevant artifacts, pointing to limitations in artifact quality, generalization, or application. Evaluating self-improving agents requires measuring not only what they can do, but how effectively they become better through experience.
Read the full abstract

All authors Harman Singh, Anton Bakhtin, Rulin Shao, Gabriel Synnaeve, Ilia Kulikov, Rob Fergus, Sanjeev Arora, Kurt Keutzer, Jason Weston, Anuj Mahajan, Anirudh Goyal

Artificial Intelligence

Sequential Probabilistic Uncertainty Estimation for Parallel Multi-Agent Reasoning Systems

Tunyu Zhang, Zihao Zhao, Yusong Zhao et al. (+5 more)

LLM-based multi-agent systems (MAS) have attracted growing attention for improving reasoning through interaction among multiple agents. In this work, we focus on parallel multi-agent reasoning systems, where several agents solve the same problem over multiple rounds and aggregate their outputs into a final answer. Despite their strong reasoning performance, uncertainty estimation for such systems remains underexplored: the reliability of a MAS depends not only on individual generations, but also on how agents interact and evolve across rounds. We propose SAUCE (Sequential Agent Uncertainty through Consensus Evolution), a lightweight, training-free uncertainty estimator that formulates MAS uncertainty as sequential inference over a latent system-level belief. SAUCE aggregates round-level agreement and generation-uncertainty signals through a filtering-style update. Across five backbones, five benchmarks, and two MAS protocols, SAUCE improves misclassification detection, selective prediction, and calibration over a broad set of uncertainty estimation baselines, including standard log-likelihood-based methods and MAS-specific estimators.
Read the full abstract

All authors Tunyu Zhang, Zihao Zhao, Yusong Zhao, Haizhou Shi, Zhuohang Li, Haoxian Chen, Hao Wang, Dimitris N. Metaxas

Artificial Intelligence

Humanize: Judgement Engineering for Agentic Coding

Sihao Liu, Ligeng Zhu, Zijian Zhang et al. (+6 more)

Agentic coding makes code generation cheap, but reliable completion remains difficult: the agent that writes the code is a weak judge of whether it is done. We present Humanize, a multi-agent orchestration workflow for agentic coding built around judgement engineering: explicit, mechanically enforced decisions at the boundaries between planning, implementation, review, and learning. A human approves a plan contract, a builder agent implements it in rounds, and a reviewer agent from another vendor decides completion; deterministic hooks, not a model, route work between these roles and enforce 72 mechanical gates. Viewed as a Markov chain over repository states, alternating builder and reviewer samples jointly from two models, so a defect survives only if both miss it. We study Humanize through its deployment, 118 public postmortems of real loops, and its applications. Over 68 versions in 108 days, it gathered 1,468 GitHub stars. Applications include a 567-file gem5 build-system migration under upstream review; Kernel Design Agents, which extend the loop with a kernel knowledge base and profiling feedback and placed in the top three of all three Full-Agent tracks of the MLSys 2026 FlashInfer contest; and, through Humanize Olympiad Agents (HOA), full scores in IOI 2026, IMO 2026, IPhO 2026, and IBO 2024, 418.5/437 in IChO 2026 (gold-medal). Humanize also achieves 672/672 on PutnamBench and ranks first (251/303) on Lean-Eval's leaderboard even competiting with professional mathematicians. The postmortems show that independent review catches unsupported builder claims, but stopping remains a key weakness. In reports that separate rounds by phase, two thirds of rounds occurred after implementation was accepted. This evidence is observational, not a controlled comparison of workflows.
Read the full abstract

All authors Sihao Liu, Ligeng Zhu, Zijian Zhang, Dongyun Zou, Zhengyang Zhang, Changye Li, Song Bian, Song Han, Tony Nowatzki

Artificial Intelligence · Cross-listed from cs.CR

Contextualization of Third-Party Cloud Security Findings

Leon Goldberg and Gal Engelberg

Finding severity is the main driver of how security teams prioritize remediation. For third-party cloud security findings, that severity is static: the rule that raised the finding assigns it before the rule meets any environment, so it reflects the risk of the condition in general rather than the risk the finding poses to the concrete environment where it lives. Scoring standards define where environment-specific context belongs. How far that context changes finding severities in production, where the deciding evidence lies, and whether it holds against the live environment have not been measured. We address this gap with contextualization, re-deriving each finding's severity from evidence in the environment where the finding lives. A deep research agent over a precomputed cross-signal asset graph investigates each finding against the resource's state, its graph neighborhood, and other products' signals, and returns an adjusted severity with an evidence trace. We evaluate it in a production field study of 9,967 vendor HIGH findings from two commercial cloud security platforms across eight real production environments, on three criteria: the faithfulness of the facts behind each verdict to the live environment, the dependence of each decision on context beyond the flagged resource, and the regularity of the reasoning. Three in four findings are re-graded, mostly downward, and the same rule often moves in opposite directions inside a single environment. About half of the decisive evidence lies beyond the flagged resource, and read-only probes of live infrastructure confirm the decisive fact for 99.4% of decided findings.
Read the full abstract

Artificial Intelligence · Cross-listed from cs.IR

Trustworthy Domain-Specific AI for Structured Knowledge Retrieval and Reasoning

Ryan C. Barron

This dissertation presents a scalable architecture for transforming unstructured, domain-specific text into structured knowledge for retrieval and reasoning. It integrates semi-automatic corpus curation, semantic structuring, retrieval, and inference into an interpretable pipeline. The research introduces Binary Bleed, an adapted binary search method that reduces low-rank search complexity for Non-negative Matrix Factorization (NMF), and Hierarchical NMF with automatic latent feature selection (HNMFk), a depth-adaptive topic modeling method that produces interpretable taxonomies guided by subject matter experts. These representations populate a typed Knowledge Graph and a semantically aligned Vector Store containing extracted latent features, synchronized through an event-driven substrate. Tensor-Structured Retrieval-Augmented Generation (T-SRAG) dynamically routes queries across retrieval paths. Contrastive alignment maps document and query embeddings to hierarchical topic structures to improve semantic fidelity and reduce hallucinations. Beyond retrieval, tensor-based link prediction identifies and completes missing links in the Knowledge Graph, supporting inference grounded in citation structure. Applications across cybersecurity, law, materials science, and healthcare demonstrate improvements in retrieval precision, early trend detection, hypothesis generation, and hallucination mitigation. The dissertation provides a deployable, modular foundation for trustworthy, domain-specific AI systems that retrieve and reason over structured knowledge.
Read the full abstract

Artificial Intelligence · Cross-listed from eess.IV

Hybrid++: The Bridge between PDE Models and Deep Learning for Gamma Noise Removal

Mahipal Jetta and Sujato Dutta

Multiplicative gamma noise is one of the dominant noise factors in Synthetic Aperture Radar (SAR) and medical ultrasound images. They are dependent on pixel level noises due to which they are highly varying across the image and harder to handle as compared to additive noise. The denoising methods to address this noise currently include classical partial differential equation (PDE) methods which are good for interpretation but lack the restoration ability as compared to the state-of-the-art models while deep convolutional networks like DnCNN achieve high performance but at the cost of transparency due to which practitioners are skeptical to use them in high-risk critical fields like medicine. This paper presents Hybrid++, a novel trainable nonlinear reaction-diffusion architecture that addresses both the concerns - staying interpretable while offering performance close to huge black-box models. It combines a fully learnable PDE initialization with a 3-stage reaction-diffusion network having 64-channel multiscale filter banks, 4-layer Squeeze-and-Excitation attention-based influence functions and a 64-dimensional noise level embedding. It uses a two-phase training strategy, stage-wise optimization followed by joint end-to-end refinement which enables co-adaptation of all learnable parameters. On the FoE benchmark, Hybrid++ substantially improves over classical PDE, BM3D and the original TNRD baselines. In the severe-noise setting L=1, it comes within 0.23 dB PSNR of a separately trained DnCNN while using only about 8% of its parameters. We therefore position Hybrid++ not as a universal state-of-the-art image restoration backbone, but as a compact, physically structured reaction-diffusion model for multiplicative gamma noise.
Read the full abstract