Artificial Intelligence · Cross-listed from cs.LG
Proteus: Incremental Memory Activation for Long-Context Sequence Modeling
Reza Bayat, Ali Behrouz, Vahab Mirrokni et al. (+1 more)
The quadratic cost of
attention-based sequence models for long contexts has motivated a growing line of research on
memory-based models that can compress context into a compact state. However, most existing
memory models expose a static
memory throughout the entire sequence. Because early
tokens face no compression pressure, they occupy too many degrees of freedom and "pollute" the
memory state, leaving little capacity for later context and increasing interference between what is stored and what arrives next. We study a new paradigm of incremental
memory activation, where the effective capacity of
memory is progressively expanded as the context grows. Imposing an early bottleneck forces the model to compress history more effectively, while unlocking fresh capacity over time reduces interference and improves retention of later context. We instantiate this paradigm in Proteus, a straightforward mechanism that can be incorporated into a broad class of neural
memory architectures at no additional cost. We apply Proteus to state-of-the-art models, including SWLA, Comba, Titans, and Hope-
Attention, and observe consistent improvements on standard
language modeling and
reasoning, as well as on long-context
retrieval and understanding, with gains that grow at longer context lengths. Overall, our results show that static
memory is suboptimal and that scheduling effective capacity is a simple and broadly applicable tool for sequence modeling.
Read the full abstract ▾
All authors Reza Bayat, Ali Behrouz, Vahab Mirrokni, Aaron Courville
Artificial Intelligence · Cross-listed from cs.RO
HAF: Adapting Generalist VLAs to Humanoid Whole-Body Loco-manipulation via Hierarchical Action Flow and Spectral Latent RL
Langzhe Gu, Chengkai Hou, Meng Li et al. (+14 more)
Humanoid robots hold great promise as general-purpose
agents in human-centered environments, yet generalist
vision-
language-action (VLA) foundation models are not readily applicable to
humanoid whole-body loco-manipulation. The high dimensionality and interdependence of
humanoid motions make it challenging for conventional single-stage VLA architectures to coordinate locomotion, waist posture, and dual-arm manipulation effectively. Moreover, policies trained through offline behavior cloning can remain suboptimal during real-world deployment. Although online reinforcement learning can refine policies through real-world interaction, directly tuning large VLA backbones demands excessive computation and may introduce safety risks during real-
robot exploration. To address these bottlenecks, we introduce HAF (
Humanoid Adaptation Framework), a two-part framework consisting of HAF-VLA and HAF-Steer that transfers off-the-shelf generalist VLA foundation models to
humanoid whole-body loco-manipulation. HAF-VLA is a hierarchical action-flow generator built on a pretrained flow-matching VLA. It splits full-body action denoising into three sequential stages with stage
embeddings and cross-stage KV caches that retain kinematic dependencies, avoiding incoherent whole-body actions from one-shot generation. On top of the frozen HAF-VLA, HAF-Steer is a latent offline-to-online RL pipeline that leverages flow-matching invertibility and DCT-based dimensionality reduction to restrict RL optimization to a compact noise subspace and train a regularized SAC policy. This avoids updating the large VLA backbone and enables efficient real-world policy refinement. Evaluated on seven real-world
humanoid loco-manipulation tasks, HAF surpasses vanilla single-stage VLA baselines and improves whole-body coordination and task performance. Project website: https://grange007.github.io/HAF .
Read the full abstract ▾
All authors Langzhe Gu, Chengkai Hou, Meng Li, Xinhua Wang, Jiaming Liu, Xinyuan Lv, Bowei Zhang, Shuanghao Bai, Guangrun Li, Jingyang He, Gaole Dai, Ziluo Ding, Zhiyuan Xu, Kuan Cheng, Jian Tang, Zhengping Che, Shanghang Zhang
Artificial Intelligence · Cross-listed from cs.CL
Model Hypnosis: Strong control of AI via additive subliminal effects
Enric Boix-Adsera and Benedict Tessler
We demonstrate that AI models are broadly susceptible to a phenomenon we call model hypnosis, in which individually weak and seemingly irrelevant cues in the
prompt can be systematically combined to strongly control model behavior. Model hypnosis occurs across model families and scales, including in frontier
reasoning models, and hypnotic
prompts can transfer between models. Because the model is controlled by inconspicuous textual choices, such as paraphrases and typos, model hypnosis presents new challenges and avenues for AI safety, and is a major hurdle for AI interpretability.
Read the full abstract ▾
Artificial Intelligence
From Answers to Policies: Efficient In-Context Learning System through Emulating Expert Investigation
Minh-Ha Nguyen, Ngoc-Ngo Quang Tran, Thuy Dung Nguyen et al. (+1 more)
Pretrained large
language models offer a practical foundation for learning useful behavior from few task-specific examples. We argue that current
prompt and context optimization methods underuse the extensive knowledge and
reasoning capabilities of trillion-parameter models. These capabilities can make adaptation more sample-efficient, more compute efficient and at no performance loss when organized around how human experts investigate failures. We formalize Policy Iteration with Human Feedback (PIHF), which makes this implicit procedure explicit for LLM
agents to execute, and build its automated implementation, PIHF-MCP. Initialized from clinician feedback on rare-disease diagnosis, PIHF-MCP supplies the expert procedure, testing tools, review and persistent inquiry records to develop reusable task policies. Across general
reasoning benchmarks (BIG-Bench Extra Hard, HoVer and LiveBench-Math), PIHF-MCP improved performance of the baseline model by 16.9, 22.2 and 4.7 percentage points, respectively. With a matched baseline model, development used about 1/5 of the labelled examples and 4% of the task rollouts reported by a previous SOTA in-context optimizer, making it about 9 times faster and 3 times cheaper at comparable or higher scores. In a low-data rare-disease diagnosis setting, policies developed from previous SOTA
prompt optimizers trailed a previously published PIHF-developed system on every held-out cohort (on average 16 percentage points). These findings support a route to more efficient inference-time scaling: PIHF-MCP develops reusable policies from a few examples that improve performance on unseen cases and across models. Because each policy comes from an explicit, recorded investigation, the process also keeps humans in the loop and enables ownership and learning, making it well suited to high-stakes decisions.
Read the full abstract ▾
All authors Minh-Ha Nguyen, Ngoc-Ngo Quang Tran, Thuy Dung Nguyen, Cathy Shyr
Artificial Intelligence · Cross-listed from cs.LG
CaliBench: Are the Stochastic Dynamics of Video World Models Physically Calibrated?
Jonathan Sadeghi, Jenny Seidenschwarz, Jesse Allardice et al. (+3 more)
Video world models approximate the
stochastic distribution of physical outcomes through generative sampling, but existing
benchmarks score individual generations or compare distributions coarsely over a whole
dataset, leaving the fine-grained aleatoric uncertainty of specific phenomena untested. We introduce CaliBench, which scores outcomes in a physically interpretable discrete space - a bin index, a die face, a suit, a colour - rather than a learned feature space such as in FID, so the distance from a known reference distribution is measured directly. We curate outcome spaces whose reference is known in closed form (binomial Galton boards, Bernoulli forks, uniform dice/cards/lottery, a skewed European-roulette colour), enabling an exact calibration test. We decompose performance into two orthogonal axes that a single accuracy metric conflates: scorability, the fraction of generations yielding a scoreable outcome, and calibration, the total variation distance from the reference on that sample. A chi-squared test assesses significance; as calibration is its null hypothesis it can evidence only miscalibration, and at N=32 per
cell detects only large deviations. We apply it to nine scenes and six image-to-
video models (WAN-2.7, SeeDance-2.0, HappyHorse-1.0, Veo 3.1, Runway Gen-4.5, Cosmos3-Super), 32 generations each. Models consistently concentrate probability mass on a few outcomes rather than reproducing the reference. Most scene-model combinations are significantly miscalibrated, in the extreme collapsing to one outcome, as Veo 3.1 does on dice. On roulette, generations often leave the ball ambiguously placed, giving several models low scorability. Performance varies by scene: no model dominates all nine. We release the protocol and a metric (mean normalised total variation, mnTV) for comparing new models against our results.
Read the full abstract ▾
All authors Jonathan Sadeghi, Jenny Seidenschwarz, Jesse Allardice, Sirish Srinivasan, Benjamin Graham, Jeffrey Hawke
Artificial Intelligence
Quipu: A Governed Bitemporal Knowledge Graph Store
Steve Brown
Agents now write knowledge graphs, but knowledge-graph stores still carry defaults set when humans curated them: accept writes now and clean later, keep one time axis or none, treat every writer's facts as equally trustworthy, and leave governance to dashboards and middleware. These four defaults are individually convenient and jointly untenable under
agent workloads. We present Quipu, an embeddable store that inverts all four: no fact enters except through a gate whose predicates evaluate the pending post-state; data, trust labels, verdicts, and the rules themselves are bitemporal; named graphs are the unit of authority and trust, composed under a
lattice whose one invariant is that composition never widens; and the governance specification $\Sigma$, the trace, and signed verdicts are facts in the store they govern, making the audit $T \models \Sigma$ a query. We evaluate with Census, a deterministic multi-writer lifecycle whose single seeded run scores every research question against planted ground truth: the gated store ends with 0 of 6 planted defects versus 6 of 6 ungated; all 7 composition probes uphold the
lattice contract; 50 of 50 satisfied verdicts re-derive faithfully as of their instant while all 50 would be misreported under a latest-only rule set; and the SARC reference checker agrees with the in-store audit verdict-for-verdict, differing only on coverage semantics. A recorded trace from a governed writer surfaces a live enforcement gap the audit names with its remediation. On DEMM-Bench, an external decision-evidence sufficiency
benchmark, a content-only reading of the exported records answers all 512 property-level governance questions correctly with zero overclaim under all eight degradation conditions, while container-presence baselines overclaim on up to 87.5% of them – and the run surfaced, and led us to close, a gap in what a denial's verdict attests.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.RO
Breaking Planner Integrity Boundary: Enviroment State-Text Injection Attack on LLM-Driven Embodied Agents
Jiawei Liu, Jiacheng Guo, Tian Zhang et al. (+7 more)
Large
language model (LLM)-driven
embodied agents rely on environment states to interpret scenes, generate high-level plans, and drive physical execution, making planner-visible state representations a critical security boundary. Existing attacks primarily manipulate user instructions,
prompt contexts, model behavior, or perceptual inputs, while paying limited
attention to whether environment-state text itself can serve as deceptive task evidence and propagate beyond planning to affect execution outcomes. Because
embodied tasks are constrained by entity grounding, action preconditions, spatial relations, and environmental constraints, planning deviation alone does not guarantee
adversarial execution. To address this gap, we investigate environment-state text as an independent attack surface and present the first closed-loop Environment State-Text Injection (ESTI) attack for LLM-driven
embodied agents. Without modifying the original user instruction, model parameters, or executor, ESTI reformulates an
adversarial objective as false state evidence compatible with the current environment and influences planning and execution through object properties, spatial relations, affordances, task-stage rules, and execution feedback. We further develop ESTI-Bench to evaluate attack propagation across the planning-to-execution closed loop and compare ESTI with Vanilla IPI, EIRAD, and BADROBOT across ProgPrompt/VirtualHome, VoxPoser/RLBench, and AI2-THOR/iTHOR. ESTI consistently outperforms existing baselines, improving planning-level and execution-level attack success rates by up to 89.32% and 43.69%, respectively. Further analysis shows that grounding, consistency, and executability jointly determine whether manipulated state evidence can propagate through the
embodied closed loop and produce verifiable environmental changes.
Read the full abstract ▾
All authors Jiawei Liu, Jiacheng Guo, Tian Zhang, Yiwei Xu, Juan Wang, Jinlin Fan, Bowen Xiao, Chi Guo, Keyan Guo, Hongxin Hu
Artificial Intelligence · Cross-listed from cs.CV
Diagnosing Dense Same-Class Attribute Misbinding in Large Vision-Language Models
Yuanzhi Xu, Qian Gao, Jun Fan et al. (+4 more)
Large
vision-
language models can recognize the objects and attributes in a crowded scene yet assign an attribute to the wrong same-class instance. Generic visual-question-answering accuracy marks the response as wrong, while object-
hallucination metrics may regard both the object and attribute as image-supported; neither reveals the transfer. This study formalizes this blind spot as Dense Same-Class Attribute Misbinding (DSCAM) and presents InstaBind-Lite, a controlled
benchmark that makes it directly measurable. Its 524 images contain 529 curated groups of 3-6 same-class entities, 1773 boxed instances, ordered neighbors, distinguishable color-like attributes, and four complementary question levels, yielding 9580 deterministically evaluated questions. Unlike existing protocols, source-instance annotations separate unsupported generation and recognition failure from an attribute copied from another visible entity. Binding-specific metrics further quantify transfer frequency, adjacency, ordinal distance, and intervention effects. Across five open-source and two commercial/API models, the open-source systems average 19.84% Misbinding Rate and the API systems 7.55%; these errors are hidden by aggregate accuracy. Among identifiable transfers, 80.70% and 81.51%, respectively, originate from adjacent instances. Localization and instance-first interventions help selected models but are not universal remedies. InstaBind-Lite therefore turns previously undifferentiated wrong answers into source-identifiable failure categories and tests a reliability dimension that conventional
benchmarks cannot determine: whether a model knows not only what is visible, but which instance owns each attribute.
Read the full abstract ▾
All authors Yuanzhi Xu, Qian Gao, Jun Fan, Guohui Ding, Zhenyu Yang, Yuteng Xiao, Sixue Lin
Artificial Intelligence
Cross-Sign Language Transfer Learning Using Domain Adaptation with Multi-scale Temporal Alignment
Keren Artiaga, Yang Li, Ercan Engin Kuruoglu et al. (+1 more)
Sign
language serves as a vital means of communication for individuals with hearing impairments, yet recognition resources for the over 100 distinct sign
languages are severely lacking. In response, we present our work on sign
language recognition using transfer learning and the domain adaptation method TA3N, which utilizes the Temporal Relational Network (TRN) module for aligning multi-scale temporal relations. Our findings highlight the superior performance of Domain Adaptation to neural network-based transfer learning, particularly in improving recognition of American Sign
Language (ASL). Our research also identifies the effectiveness of aligning shorter-term temporal features between source and target domains. In addition to using RGB, we conducted experiments using Optical Flow mode for the sign
language samples, ultimately determining that RGB outperforms Optical Flow in the majority of cases. Our work aims to improve accessibility and communication for individuals who rely on sign
language as their primary mode of communication.
Read the full abstract ▾
All authors Keren Artiaga, Yang Li, Ercan Engin Kuruoglu, Wai Kin Chan
Artificial Intelligence
When Agents Coordinate: Measuring Coordination in Multi-Agent AI Coding
Giuseppe Destefanis and Tomaso Aste
We study how teams of AI coding
agents coordinate while solving programming tasks. Current evaluations usually report whether the
agents complete the task and how much the run costs, leaving the coordination inside the team largely unmeasured. We introduce an instrument to measure this coordination. Each run is represented as a temporal network in which
agents and files are nodes, and messages, file writes, and file reads are timestamped directed edges with an associated cost. We apply this instrument to 1902 runs, each evaluated with a fixed test suite, across configurations that vary the team size, the team structure, and the file policy. The resulting networks show how coordination changes as teams grow and as the work changes. Direct messaging initially increases close to quadratically with the number of
agents, with much of this growth coming from an early round of introductions. As the teams grow further, this increase levels off in the largest teams we study, where
agents increasingly communicate through broadcast messages. The task also shapes the network that emerges. Work built around a shared specification produces dense, highly connected teams, while pipeline tasks produce sparse networks organised around local interfaces. Shared files can replace repeated 1-to-1 communication, cutting output
tokens by about 42% at eight
agents on message-heavy work, while adding overhead when files already carry the coordination. Naming one
agent as coordinator creates no communication hub and provides no reliable improvement in success. We also observe an unprompted tendency for
agents to seek out hidden grading material. We repeat the key experimental conditions in a sealed environment, replacing the hidden material with marked placeholder files. Across 244 additional runs,
agents still reach for it in four fifths of runs, while the coordinator and file-channel findings reproduce.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
ClawGym II: Exploring Black-Box RL on Agent Harness
Huatong Song, Fei Bai, Ming Yang et al. (+17 more)
Agent harnesses have substantially improved performance on long-horizon tasks by coordinating
agent interactions with the environment. However, reinforcement learning through complex harnesses remains largely unexplored, as scaling such training to long-horizon
agent tasks introduces fundamental challenges. In this work, we present a unified black-box RL framework for stable and scalable optimization of general
agents through complex harnesses. Concretely, we first build a sandbox-based execution infrastructure that isolates task environments and harnesses within temporary sandboxes for large-scale concurrent rollouts. We then decouple policy optimization from opaque harness execution and place a serving proxy at the model boundary to capture model calls. To reconstruct multi-turn trajectories and improve training efficiency, we organize the captured calls into prefix trees and further adapt both critic-based PPO and critic-free GRPO to optimize over the recovered tree structure. Meanwhile, we maintain training-inference consistency throughout the optimization process. Finally, we introduce mix-harness training, allowing a single model to be jointly optimized by heterogeneous harnesses. With Qwen3-30A3B, black-box RL improves Pass@1 on ClawGym-Bench by 9.98 and 14.81 points through OpenClaw and Claude Code, respectively, while remaining stable over 200-400 optimization steps. Moreover, the framework yields consistent gains on more challenging tasks such as JobBench and OfficeQA. Overall, our framework enables effective, stable, and scalable optimization of general
agents through black-box harnesses, supporting unified training across heterogeneous execution systems.
Read the full abstract ▾
All authors Huatong Song, Fei Bai, Ming Yang, Renyuan Li, Jia Deng, Jujie He, Zhange Zhang, Daixuan Cheng, Yan Xing, Qi Yun, Xuxing Chen, Danyang Li, Feng Chang, Chuan Hao, Ran Tao, Jian Yang, Bryan Dai, Wayne Xin Zhao, Mingjie Tang, Ji-Rong Wen
Artificial Intelligence · Cross-listed from cs.IR
UniDot: A Unified Network for Sequence Modeling and Feature Interaction in Large-scale Recommendation
Rongcheng Lin, Yan Sun, Jamey Zhang et al. (+4 more)
Industrial recommenders rely on two model families that have evolved largely independently: feature-interaction models over multi-field user/item features, and sequential models over user-behavior histories. Production systems couple them only loosely. To unify the two, we present UniDot, a novel architecture for post-click conversion prediction built from the factorization-machine (FM) point of view: the
embedding inner product—which powers collaborative filtering and lets a recommender generalize to unseen user–item pairs—is the same primitive as
attention's query dot key scoring, so a single dot-product of
tokens can underlie both feature interaction and sequence modeling. UniDot tokenizes non-sequential fields and multi-domain behavioral sequences into one shared
token space and stacks a single macro-block in which a
token-mixing bus and a sequence-
retrieval bus (item
tokens cross-attending the histories) run in parallel and exchange state each layer through an MLP-Mixer fusion, while an FM Highway carries explicit per-layer dot-product interactions around the residual stack directly to the classifier. The sequence side is embedded once per forward pass and shared by all consumers, bounding inference latency. Trained with a dual sparse/dense (Adagrad +
Muon) optimizer, an auxiliary conversion-delay head, and multi-path mutual learning, UniDot finished as the runner-up on the Industrial track of the TAAC KDD Cup 2026.
Read the full abstract ▾
All authors Rongcheng Lin, Yan Sun, Jamey Zhang, Guanglei Xiong, Ivan Ji, Xianjie Chen, Shujian Bu
Artificial Intelligence · Cross-listed from cs.CE
Historical Backtesting for Scientific Question Discovery: A Protocol and Astronomy Pilot
Hui Mao
Systems that generate scientific research questions are evaluated today by expert scores, LLM-as-judge ratings, or curated case studies – all subjective, none falsifiable. We formalize historical backtesting as an alternative: a system generates questions from a corpus frozen at a historical cutoff, the questions are frozen before any access to later literature, and a temporally isolated future corpus then determines whether each question was subsequently answered, partially addressed, independently posed, or ignored, and whether its underlying premise was supported or refuted. The protocol is model-agnostic: any system that emits frozen questions can be scored. We release reproducible astronomy instances with temporally isolated corpora, frozen questions, auditable labels, four reference baselines, and a submission interface. Two findings result. First, evidence-structure-first generation outperforms LLM-only prompting: across a generator decomposition crossed with a four-cutoff stress test (2010-2024, 798 judged questions) whose last window postdates model training, LLM-only generation shows memorized relevance without specific foresight, while a generator using no model weights at all finds questions whose premises the future refutes in every era. Second, a seven-rater agreement study (two blinded human annotators, five judge models, 90 items) indicts the outcome taxonomy rather than the judge: two careful humans agree at kappa = 0.17, every judge model agrees with the professional annotator as well or better (0.17-0.26), and frontier models agree with one another at 0.60 – certifying an LLM judge by model-model agreement would have overstated its reliability threefold. A prospective instance – 200 questions frozen 2026-08-17, scored 2027-2030 – is released so the central claims become contamination-free tests that time itself will grade.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.RO
Mohammad Albinhassan, Yuming Feng, Alessandra Russo et al. (+1 more)
Language and
vision-
language models generate plausible
embodied plans but do not guarantee executability, as their outputs can violate environment dynamics or act on incorrectly grounded entities. We present a neurosymbolic
agent that factors long-horizon household tasks into task-directed visual exploration and constrained symbolic planning. In the first phase, a
vision-
language model and exploration harness acquire goal-relevant predicates and instance bindings from
egocentric observations and grounded interactions, producing a symbolic initial state. In the second, a PDDL transition model restricts decoding to
tokens that extend applicable actions. Monte Carlo tree search then evaluates executable continuations using a domain-independent planning heuristic. The resulting plans are executable by construction under the transition model, with transfer to the environment conditioned on correct visual grounding. On VirtualHome and ALFWorld, open 4B-27B models exceed 90% success in both environments, and our smallest
agent substantially outperforms a 27B direct visual policy in each. Constraints and search prove complementary rather than interchangeable: in ALFWorld either alone solves under a third of tasks, whereas their combination solves over 95%. The method also uses several times fewer generated
tokens than extended thinking and far fewer model-visible images than direct interaction, and residual failures localize to state acquisition rather than plan generation without any specialized training.
Read the full abstract ▾
All authors Mohammad Albinhassan, Yuming Feng, Alessandra Russo, Pranava Madhyastha
Artificial Intelligence · Cross-listed from cs.CV
Steering the Flow: Inverting Face Recognition Models via Gradient-Guided Flow Matching
Ye Lu, Shen Wang, Zhaoyang Zhang et al. (+4 more)
Model Inversion Attacks (MIAs) aim to reconstruct representative training samples of target identities from face recognition models, exposing critical security vulnerabilities. Existing methods typically rely on indirect guidance or highly
stochastic guidance, making it difficult to stably optimize generation trajectories toward target facial images. In this paper, we propose Steering Flow Model Inversion (SFMI), a novel two-stage white-box model inversion method that reformulates inversion as a trajectory-steering task. Specifically, Step I, Learning a Generic Flow Matching Prior, pre-trains a generic unconditional Flow Matching model to encode the
manifold of human faces as a robust prior. Step II, Attacking with Progressive Guidance Scheduler (PGS), injects time-dependent target-specific gradients during sampling. By backpropagating through the target model to obtain gradients from intermediate generated states, PGS progressively injects adaptive guidance signals into the vector field. This process effectively steers the current generative flow from random noise toward the high-density regions of the target class. Under an identity-disjoint cross-evaluation setting using the CelebA
dataset, SFMI achieves an ACC of 0.9248, an FID of 22.61, and an LPIPS of 0.3874 on the ArcFace target. Extensive experiments on multiple target models demonstrate that SFMI achieves competitive state-of-the-art performance in attack success and visual fidelity under the evaluated white-box protocol.
Read the full abstract ▾
All authors Ye Lu, Shen Wang, Zhaoyang Zhang, Yihan Yan, Li Liu, Runze Liu, Fanghui Sun
Artificial Intelligence
GRIP: Grounded Reasoning via Information-Restricted Premises
Lirui Teng
High-capacity encoders in
retrieval-augmented generation (RAG) can let the query dominate the latent state, leaving retrieved evidence functionally irrelevant. We call this failure mode query dominance. To address it, we introduce GRIP (Grounded
Reasoning via Information-Restricted Premises), which imposes capacity asymmetry: the decoder keeps full-dimensional access to the query, while retrieved evidence passes through a severe
stochastic bottleneck. This forces the evidence channel to encode only the residual information unavailable from the query. Across five
reasoning benchmarks, GRIP outperforms strong iterative baselines, cuts a query–latent mutual-information diagnostic by roughly 30$\times$ (14.8 $\to$ 0.47 bits), and reduces
hallucination by 73%. Residual-
alignment analysis further shows that the bottleneck output occupies subspaces less aligned with the query than baseline representations.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CR
Topological Attribution Distance (TAD): Revealing Segment-Level RAG Influence on LLM Output Geometry for Incident Log Analysis
Reza Fayyazi, Michael Zuzak and Shanchieh Jay Yang
Large
Language Models (LLMs) are increasingly being deployed in cybersecurity operations to assist cybersecurity analysts with rapid decision-making against emerging threats. However, there is a main criteria that must be met when using LLMs in cybersecurity, that is, trust in the generated outputs. As Agentic AI is integrated into operational systems, a robust evidence attribution and provenance tracking technique is essential to trace the origins of model generations. When autonomous
agents make a decision (right or wrong), the ability to trace back through the decision chain is critical, as without it, teams cannot identify which segment of the data caused the model generation. Existing methods often struggle to distinguish among complex and highly similar evidence sources, such as cyber incident logs. This reveals a key gap: current approaches do not adequately capture the holistic
geometric relationship between the retrieved evidence and the generated response for reliable evidence verification. To bridge this gap, we propose Topological Attribution Distance (TAD), inspired by
Topology, to characterize and capture the global
geometric shape of an output and its changes against its retrieved logs. In other words, if the
embeddings of a specific source log drastically changes the geometry of the model's response in the
embedding space, this suggests that such log is a critical source for the model's generated response. Therefore, TAD is powered by segment-level ablation attribution to investigate incident logs of an actual cyberattack. We demonstrate how TAD finds the most attributed logs on LLM outputs in an adaptive manner. This can provide an explainable and trustworthy tracing based on each LLM's hidden state to understand how geometrically different retrieved logs influence the model generation, and provide evidence verification in cybersecurity and Agentic-AI workflows.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CV
TRACE-Bench: Decomposing and Diagnosing Multi-Reference Image Generation
Haoran Wang, Chaofan Ma, Ran Yi et al. (+1 more)
Despite recent advances in unified
multimodal models for multi-reference image generation, existing
benchmarks remain organized around predefined task types (e.g., "subject composition"), which are ill-suited to this combinatorial setting and lead to fragmented coverage, uncontrolled complexity, and little diagnostic value. Recognizing that diverse multi-reference tasks share a common set of atomic operations, we adopt a capability-oriented perspective and formalize four operators: Anchor ($f$), Disentangle ($g$), Apply ($\oplus$), and Compose ($C$). Any multi-reference
prompt can then be represented as a compositional formula over these operators, whose structural complexity is quantified by the number of operator slots. Building on this formulation, we construct TRACE-Bench, comprising approximately 1,600 evaluation cases across slot counts 1–8, built from 631 formula templates and around 4,000 reference images spanning diverse artistic styles and real-world subjects. The formula structure directly drives an operator-aligned evaluation protocol for per-capability scoring and a diagnostic tree analysis for recursive failure localization. Evaluating 9 leading models reveals insights invisible to holistic scoring: the primary bottleneck lies in disentanglement ($g$) and attribute binding ($\oplus$) rather than scene-level composition ($C$), with even the best model scoring only 0.74 on attribute fidelity. Project page: https://amuseum-whr.github.io/TraceBench
Read the full abstract ▾
All authors Haoran Wang, Chaofan Ma, Ran Yi, Lizhuang Ma
Artificial Intelligence
LAVA: Logic-Aware Validation and Augmentation Framework for Large-Scale Financial Document Auditing
Ruoqi Shu, Xuhui Wang, Isaac Wang et al. (+2 more)
Financial document validation in production, such as payroll
auditing, tax compliance, and loan underwriting, demands exceptional accuracy, consistency, and reproducibility under strict enterprise constraints. In practice, documents arrive with heterogeneous layouts and formats, semantically rich and context-dependent content, and embedded business rules that current pipelines struggle to process reliably. We introduce LAVA (Logic-Aware Validation and Augmentation), a modular, backbone-agnostic pipeline built on
multimodal large
language models, that integrates a four-stage design: document-rule
retrieval, layout-preserving information extraction, auxiliary metadata enrichment, and auditable symbolic/arithmetic verification. LAVA supports robust rule grounding, fine-grained error attribution, and consistent, traceable end-to-end execution, capabilities essential for high-stakes deployment. Evaluated on a large real-world
benchmark with diverse financial documents and dozens of expert-curated validation rules, LAVA outperforms baselines in
hallucination control and edge-case handling while maintaining efficient
token usage, demonstrating practicality for high-volume, time-critical validation.
Read the full abstract ▾
All authors Ruoqi Shu, Xuhui Wang, Isaac Wang, Yanming Mai, Bo Wan
Artificial Intelligence · Cross-listed from cs.LG
Would this change your answer? Evaluating Explanations of LLM Behavior In The Wild with Counterfactual Experiments
Adam Karvonen, Euan Ong, Subhash Kantamneni et al. (+1 more)
Many areas of AI research, such as
language model interpretability and chain of thought faithfulness, seek to explain model behaviors. But what constitutes a "good" explanation? In this work, we evaluate explanations through the lens of counterfactual simulatability-whether the explanation is useful for predicting model behaviors on related counterfactual inputs. To this end, we introduce CHIVE (Counterfactual Hypothesis Investigation Via Edits), a novel agentic pipeline that identifies unexpected model behaviors in the wild and investigates them with counterfactual
prompt edits. This yields thousands of high-quality explanations for naturally-occurring model behaviors along with supporting counterfactual evidence. We apply CHIVE in two ways. First, we evaluate whether common LLM interpretability techniques improve an
agent's ability to predict counterfactual model behaviors. Surprisingly, we find no uplift from any of the interpretability techniques studied. Second, we use CHIVE to generate training data. We find that training models to predict outcomes of CHIVE-generated counterfactual experiments generalizes to various out-of-distribution settings. Overall, CHIVE automatically discovers explanations of naturally-occurring LLM behaviors, enabling us to evaluate and improve methods for explaining LLM behaviors.
Read the full abstract ▾
All authors Adam Karvonen, Euan Ong, Subhash Kantamneni, Samuel Marks
Artificial Intelligence · Cross-listed from cs.SE
TDD-Agent: Test-Driven Reasoning for Code Generation
Hongyue Yu, Kefan Li, Jiakun Li et al. (+4 more)
Large
Language Models (LLMs) have achieved remarkable progress in code generation, yet ensuring correctness in complex, repository-level tasks remains challenging. Existing approaches often use generated tests as static post-hoc validators, which limits their ability to guide implementation and may introduce misleading feedback when the tests themselves are incomplete or incorrect. In this paper, we introduce TDD-
Agent, which operationalizes the test-driven development paradigm for code generation. TDD-
Agent first
prompts the model to generate executable tests, encouraging it to clarify expected behaviors before implementation, and then performs iterative dual-track refinement over both the generated code and tests using execution feedback. We first isolate the effect of test-first
reasoning through a
prompt variant TDD-
prompt on LiveCodeBench, where it consistently improves upon
reasoning-based prompting baselines. Building on this finding, we evaluate the full TDD-
Agent framework on RepoEval, a repository-level
benchmark, and show that it consistently outperforms
retrieval-based and
agent-based baselines. Additional analyses show that iterative refinement improves not only code correctness but also the effectiveness of the generated tests, yielding higher pass rates, coverage, and mutation scores, suggesting that tests can serve as evolving
reasoning artifacts rather than fixed validators. Our source code is available at https://anonymous.4open.science/r/TDD-
Agent-Framework-6370/.
Read the full abstract ▾
All authors Hongyue Yu, Kefan Li, Jiakun Li, Hongzheng Chai, Yuan Yuan, Rui He, Junyi Wei
Artificial Intelligence · Cross-listed from cs.AR
GoalEvolve: From Handcrafted Algorithm Priors to Goal-Driven Evolution of Physical Design Algorithms
Haixu Liu, Lei Zhou, Yuhao Ren et al. (+2 more)
Physical design
algorithms operate within tightly coupled, multi-stage optimization flows, where stage-local gains may vanish or induce downstream degradation. Existing program-
evolution frameworks often rely on stage-local objectives or undifferentiated multi-metric feedback, which neither guarantee better final results nor identify which unmet requirement should guide the next iteration. We present GoalEvolve, a goal-driven framework that makes physical design
algorithm evolution accountable for the final quality of results (QoR) of the complete flow. Given a multi-objective QoR target region, GoalEvolve converts unmet requirements into normalized target gaps, identifies the dominant bottleneck, and uses stage-resolved checkpoint evidence to locate the responsible stage. An LLM-based Teacher then narrows the search to a relevant algorithmic decision and source region, while parallel Student
agents implement and validate hypotheses through full-flow evaluation. Local effects, optimization debt, and downstream retention are retained as mechanism evidence for subsequent
evolution. Across eight ASAP7 designs, GoalEvolve improves post-route TNS by 30.67% on average and reduces leakage and dynamic power by 21.18% and 9.42% versus default OpenROAD. Relative to commercial-tool goals, it closes 62.20% of the normalized power gap on power-dominant designs, surpasses the TNS goals on both timing-dominant designs, and closes 32.48% of the equal-weight timing-power gap on joint designs. Across all three designs evaluated against Codex goal mode under matched budgets, GoalEvolve further improves TNS by 26.46% while reducing leakage and dynamic power by 12.38% and 0.76%, respectively.
Read the full abstract ▾
All authors Haixu Liu, Lei Zhou, Yuhao Ren, Yumao Wu, Zhiang Wang
Artificial Intelligence · Cross-listed from cs.CV
Unsupervised Anomaly Detection for Image Dataset Quality Assurance in Multi-Center Breast MRI
Chiara Tappermann, Steffen Renisch, Lars Ole Schwen et al. (+3 more)
Corrupted, inconsistent, or anomalous data silently threatens the safety and reliability of medical AI. Despite growing regulatory recognition of
dataset quality assurance (QA) for high-risk medical AI, scalable automated detection remains underdeveloped. We employ unsupervised anomaly detection (AD) and out-of-distribution (OOD) detection as an automated
dataset QA mechanism for multi-center dynamic contrast-enhanced breast MRI. We build a controlled AD
benchmark of 17 realistic QA-relevant anomaly types from six public
datasets (protocol violations, processing errors, incorrect anatomical regions) and propose a taxonomy of radiological image anomalies based on human visual perception, enabling fine-grained analysis of AD failure modes. The
benchmark includes near-, medium-far-, far-OOD samples, as well as in-distribution and external normal data. Four methods are evaluated: a projection-based method extended with a domain-specific feature extractor and a novel positional encoding, a reconstruction-based approach extended to full 3D volumes with an augmented training objective, and two unmodified hybrid OOD detection methods. Medium-far- and far-OOD samples are detected reliably, whereas near-OOD samples and external normal data from unseen institutions expose method-specific differences. The 3D reconstruction-based approach best balances detection performance (AUROC: 0.936) and generalization to unseen institutions. The projection-based method with positional encoding achieves the highest overall detection performance (AUROC: 0.954). Both hybrid methods exhibit critical failure modes, confirming that methods validated for one modality or anatomy may not generalize without domain-specific adaptation. Implants and mastectomies remain an open challenge for all methods. Our results establish a foundation and practical guidance on scalable unsupervised QA in medical AI pipelines.
Read the full abstract ▾
All authors Chiara Tappermann, Steffen Renisch, Lars Ole Schwen, Hans Meine, Horst K. Hahn, Eike Petersen
Artificial Intelligence · Cross-listed from cs.CV
MIRROR: Multimodal Intelligent Radiology Reasoning and Observation Reporter
Vignesh Nagarajan and Sriram Venkatapathy
A radiologist reading a model's output faces two problems. The model returns a number and no reason, and any system that turns that number into readable prose can quietly add claims the model never made. MIRROR is a research prototype built to separate those failures. It chains a multi-label classifier, a Grad-CAM localizer that turns each positive finding into a named anatomical region, and a report writer that receives the labels, probabilities, and regions but never the image. Because the
language layer cannot see pixels, it cannot assert a finding the classifier did not make. We are precise about what that buys: a MIRROR report's findings are auditable against the probability vector, while the sentences framing them are ordinary generated text, and we show one stating a cardiothoracic ratio the system never measured. One registry holds the taxonomy, anatomy, and phrasing for chest X-ray,
brain MRI, and head CT, so adding a modality is a data change; all three are routed and tested, one is trained. On ChestMNIST that classifier reaches macro AUROC 0.729 and ranks better than chance on all 14 labels, at 1.6 to 6.8 times the precision a random ranker would get. Yet at the default 0.5 threshold it emits no positive prediction at all for 11 of them, and its excellent-looking Brier score of 0.045 sits beside the 0.047 earned by a predictor that ignores the image. The discrimination is real; the decisions are not. Under the class imbalance normal in radiology, aggregate metrics flatter models that do nothing, and should be reported against that floor.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.CL
Semantic Bandits: In-Context Exploration-Exploitation is Biased by Semantic Priors
David Eric Austin, Kaheer Suleman and Jackie Chi Kit Cheung
Large
language models (LLMs) are increasingly deployed as decision-making
agents in settings that require sophisticated environmental exploration. However, existing work has raised questions about how LLMs actually balance exploration and exploitation. Unlike classical
agents, LLM
agents engage with tasks through natural
language, exposing them to semantic information with no formal counterpart in the task structure. We introduce the semantic bandit, an extension of the multi-armed bandit setting that explicitly considers the textual labels assigned to actions, and use it to study how semantic priors — inductive biases arising from associations between
language and expected reward learned during pre-training, shape LLM exploration behaviour. We find that semantically informative action labels reduce exploration in favour of exploitation, improving performance when aligned with the reward structure and severely degrading it when misaligned. We further find that negative rewards trigger substantially more exploration than equivalent positive rewards, consistent with an expected-scale bias induced by reward conventions common in pre-training data. Overall, we argue that the use of
language to define the environment and rewards introduces unavoidable biases derived from the fact that the model is trained on word co-occurence, with implications for the reliability and robustness of LLM
agents in real-world decision-making settings.
Read the full abstract ▾
Artificial Intelligence · Cross-listed from cs.LG
Learning to Unlearn: Machine Unlearning via Learning the Unlearning Behaviors
Hang Zhang, Kaifeng Zhang, Yixiao Ma et al. (+3 more)
Various machine unlearning techniques have been developed in response to
privacy legislation requirements, enabling individuals to exercise their legal right to have their data $D_f$ removed from a machine learning model. This process is typically accomplished via the use of an unlearning function denoted as $U$. Existing methods focus on designing an intricate $U$ to unlearn $D_f \subset D$ from a previous model $A(D)$, so that the unlearned model performs as closely as possible to the retrained model $A(D \setminus D_f)$. However, these methods often suffer from high computational costs when dealing with massive training data, as the complex structures of $U$ become a bottleneck even for models with fewer parameters. Inspired by Learning to Optimize, we introduce the first learning-based model-agnostic approach, Learning-to-UnLearn (L2UL). Our core insight is to shift from manually designing $U$ to learning the unlearning behaviors from a distribution perspective, thereby acquiring a simple and efficient $U$ via learning. Our experimental results demonstrate that the accuracy achieved by L2UL is comparable to that of retraining while exhibiting impressive efficiency, particularly in data-intensive scenarios. Furthermore, we validate the performance and scalability of our method on larger models ResNet.
Read the full abstract ▾
All authors Hang Zhang, Kaifeng Zhang, Yixiao Ma, Weijie Xu, Ye Zhu, Kai Ming Ting
Artificial Intelligence
MAE I Trust Myself? Self-Evaluating VLA Action Generation with Markov Attention Entropy
Aniri, Chen Yilin, Jinhe Bi et al. (+7 more)
Vision-
Language-Action models (VLAs) integrate visual perception,
language instruction, and action generation into end-to-end policies across heterogeneous architectures. However, enabling VLAs to self-evaluate their action generation reliability without external supervision remains a major challenge. Existing methods either rely on expert annotations or estimate uncertainty only from output statistics, largely ignoring internal signals. In this work, we observe that internal visual modality
entropy exhibits consistent distinctions between successful and failed tasks across heterogeneous VLAs. Although VLAs' architectures differ in their action generation, we show that they share a common latent action generation abstraction evolving under visual perception,
language instruction, and State Input, which we formulate as a Conditional Generative Markov Chain. Based on this formulation, we propose MAE (Markov
Attention Entropy), a self-evaluation framework that directly converts internal
attention signals into architecture-aware reliability scores, and introduce LIBERO-Reflect, a 4,000-episode
benchmark combining 2,000 standard episodes and 2,000 challenging episodes across four subsets. Extensive experiments across heterogeneous VLA architectures and diverse scenarios show that MAE consistently outperforms state-of-the-art baselines on AUPR, AUROC, and FPR@95.
Read the full abstract ▾
All authors Aniri, Chen Yilin, Jinhe Bi, Zengjie Jin, Yujun Wang, Yijun Tian, Volker Tresp, Fei Shen, Tat-Seng Chua, Yunpu Ma
Artificial Intelligence · Cross-listed from cs.LG
UniTAC: Universal Task-Aware Compression via Weighted Distortion Measures
Homa Esfahanizadeh, Matin Mortaheb, Adeel Mahmood et al. (+2 more)
Lossy compression is conventionally driven by a task-agnostic distortion (e.g., MSE or MS-SSIM), yet in many emerging applications the receiver cares not about uniform fidelity but about a downstream task whose relevant content varies across the signal and evolves over time. We formulate task-aware compression as a weighted rate-distortion problem, in which a single codec is driven by a separable, per-component weighted distortion whose weights encode task importance and may depend on the source. We introduce task consistency, i.e., that minimizing the weighted distortion also minimizes the true task loss, and characterize when it holds: for linear tasks, the task loss admits a weighted-MSE form with signal-independent weights under suitable cross-term conditions, while for nonlinear tasks, an integrated-gradients analysis motivates separable task-aware weights. We show how task
symmetry and irrelevance further constrain the admissible weights. Guided by this theory, we realize the weight-conditioned code in a single learned
Vision Transformer (ViT) codec whose
token-level
attention natively consumes a per-component importance vector, so one fixed backbone is re-targeted at runtime, from universal (task-agnostic) to task-specialized operation, purely by swapping the injected weights, without retraining, while producing a single human-viewable reconstruction steered to the active task. On downstream face-analysis tasks, a single model reaches 91.4% accuracy at 0.034 bpp on a localized task, within 1.9% of a task-specific codec (93.3%) and well above a universal codec (76.9%). Such task-adaptive compression suits bandwidth-constrained perception systems, e.g., in Physical AI, where the active task drifts and per-task retraining is infeasible.
Read the full abstract ▾
All authors Homa Esfahanizadeh, Matin Mortaheb, Adeel Mahmood, Jinfeng Du, Harish Viswanathan
Artificial Intelligence · Cross-listed from cs.CV
Bridging the Gap between Labeled and Unlabeled Data via Unified Flow with Feature Memory Bank
Shanwen Wang, Xin Sun, Danfeng Hong et al. (+2 more)
Although semi-supervised semantic
segmentation ($\text{S}^4$) utilizes abundant unlabeled data to reduce manual labeling burdens, independent training of labeled and unlabeled data causes the former to dominate, which severely degrades pseudo-label quality. To address this challenges, we propose a novel remote sensing (RS) $\text{S}^4$ method via unified flow with feature
memory bank (UFFM). Specifically, UFFM comprises two key innovations: unified flow (UF) and feature
memory bank (FMB). The UF is a new training flow that generates less biased pseudo-labels by combining an external visual foundation model (VFM) with an RS domain teacher, and jointly optimizes labeled and pseudo-labeled data under a unified training objective. The FMB is a novel
memory module for $\text{S}^4$ that dynamically updates class-specific features during training and reduces the feature discrepancy between labeled and unlabeled data through class-feature
alignment. To verify the effectiveness of our model, we conduct extensive experiments on RS
datasets. The experimental results show the superiority of our method over SOTA $\text{S}^4$ methods. Moreover, the results demonstrate the effectiveness of our contributions in bridging the optimization and feature representation gap between labeled and unlabeled data. Our code is released at \href{https://github.com/wangshanwen001/RS-UFFM}{https://github.com/wangshanwen001/RS-UFFM}.
Read the full abstract ▾
All authors Shanwen Wang, Xin Sun, Danfeng Hong, Junyu Dong, Patrick Le Callet
Artificial Intelligence
Chronocooked: A Benchmark for Interval Timing in Reinforcement Learning Agents
Amrapali Pednekar, Alvaro Garrido-Perez, Yara Khaluf et al. (+1 more)
Interval timing is extensively studied as an important aspect of human behaviour. As artificial
agents are increasingly designed to function alongside humans, their interval timing abilities also needs to be studied. However, research in this area remains limited and scattered. This paper presents Chronocooked, a reinforcement learning (RL)
benchmark environment that enables a systematic study of interval timing abilities in RL
agents. Inspired by Overcooked, the suite comprises cooking scenarios involving interval timing tasks drawn from the psychology literature. The tasks and reward functions are designed such that temporal information is unobserved but critical for optimal performance. The environment is intentionally kept simple to enable controlled experiments and support biologically plausible models. Each task is accompanied by evaluation metrics to study different aspects of interval timing in RL
agents, namely, task performance, human-like timing and scalability. We report baselines using a non-recurrent architecture (CNN), a recurrent architecture (LSTM), and a biologically inspired recurrent architecture (CTRNN). The baseline model analysis shows that, although RL
agents can successfully perform time-dependent tasks, they do not necessarily process and perceive time in the same way as humans. Understanding these differences is important for anticipating their impact on human-
robot interactions (HRI).
Read the full abstract ▾
All authors Amrapali Pednekar, Alvaro Garrido-Perez, Yara Khaluf, Pieter Simoens