# Research Papers — 2026-09 · bundle page 120 of 266 > Full-text markdown of papers #5951–6000 of 13298 published in 2026-09, newest first. AI-readable mirrors of the full 13298-paper 2026-09 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-09/page/120 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/119.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/121.md) ### Jev-Mem: System-One-Controlled Agentic Memory for Efficient AI Agents Authors: Dongming Jiang, Yi Li, Bingzhe Li Published: 2026-09-21 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.23986v1 Agentic memory is becoming essential for long-horizon AI agents, yet many existing systems rely on autoregressive LLMs to control how memories are organized, retrieved, and used, placing expensive generation on the critical path of memory operations. We introduce \textbf{\method}, a new agentic memory architecture inspired by System-One/System-Two cognition. System One captures fast, lightweight decision-making, whereas System Two performs slower, deliberative reasoning. Jev-Mem brings this division of labor to agentic memory through a dedicated System-One control plane, a structured multi-relational memory plane, and a System-Two reasoning plane. The System-One controller governs memory typing and relational organization during construction, and dynamically performs query routing, retrieval-budget allocation, graph traversal, candidate scoring, and adaptive stopping during retrieval. System Two is invoked only for complex reasoning and answer synthesis. This design improves both memory effectiveness and system efficiency: on LoCoMo Jev-Mem achieves an overall LLM-as-a-Judge score of 0.777, an 11.0\% relative improvement over the strongest baseline, while reducing memory construction time to 158\,s, a 6.6$\times$ speedup over the fastest competing memory system, and lowering average query latency to 0.93\,s, a 36.7\% reduction. --- ### Evaluating the Generalization of Neuroimaging Foundation Models on African Brain MRI Authors: Oluwatobi Iyanuoluwa Akinmuleya, Olatokun Shamsudeen Akano, Samuel Danquah Ankapong, Olamide Lawal, Toufiq Musah Published: 2026-09-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.23983v1 Neuroimaging foundation models pretrained on large, predominantly western cohorts are increasingly proposed as general-purpose backbones for brain MRI analysis. Yet, their ability to generalize to underrepresented clinical populations remains largely untested. We evaluate four recent foundation models (BrainIAC, Neuro-JEPA, NeuroVFM, and Primus) on a three-way diagnostic classification task (Control, Dementia, Parkinson's disease) using a cohort of 88 subjects from a Nigerian clinical brain MRI dataset, across four modality configurations (T1w, T2w, T1w+T2w, FLAIR), and compare against an end-to-end trained ViT3D baseline. The frozen backbones collapse to majority-class predictions, while Neuro-JEPA on FLAIR shows modest but still limited discrimination. In contrast, the end-to-end trained ViT3D achieves higher accuracy and MCC on every task (up to 53.4% accuracy, MCC=0.27) and is the only model with non-trivial recall. Our findings suggest that these frozen neuroimaging foundation models are insufficient for fine-grained diagnostic classification in small, non-western clinical cohorts, motivating parameter-efficient adaptation and broader multi-site external validation for equitable deployment in global health settings. --- ### MobileCybench: Evaluating Agent Vulnerability Discovery via Executable Probes Authors: Andy K. Zhang, Ava Huang, Joey Ji, Wai Han, Thomas Qin, Nardos Demilew, Michael Tian-Yue Liu, Brian Song, Riya Dulepet, Brian Wang, Kyleen Liao, Cuiyuanxiu Chen et al. Published: 2026-09-21 Categories: cs.CR, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.23980v1 AI agents now report vulnerabilities faster than maintainers can review them. Reports often depend on security properties specific to the application, and require considerable human labor to process. To mitigate this, we introduce a framework for evaluating vulnerability reports via probes, executable checks of security properties. A reported exploit is evaluated by replaying it against the application and running the probes: a triggered probe indicates both that the exploit succeeded and which security property it violated. As a probe encodes a security property rather than a known vulnerability, it can detect vulnerabilities that were not known when the probe was written. We instantiate the framework as MobileCybench, a benchmark for vulnerability discovery by AI agents in 13 Android applications, with 495 probes written and reviewed by the authors. We evaluate 5 coding agents (OpenCode with GPT-5.5, GPT-5.6-Sol, and GLM-5.2; Claude Code with Opus 4.8 and Opus 5) under 4 settings: as a malicious app on the victim's device or as a remote attacker with a low-privilege account, each with either only an obfuscated APK or access to the application's source code. Given only the obfuscated APK, the top agent, OpenCode with GPT-5.6-Sol, triggers probes in 53.8% of applications in the malicious-app setting and 16.7% in the remote-attacker setting. With source code, the trigger rate across all agents and both attack settings increases from 28.8% to 32.8%. Building and running the benchmark surfaced 23 previously unreported vulnerabilities, the majority of which have been confirmed by maintainers. --- ### AURA: Uncertainty-Routed Activation Editing for Acoustic Grounding in Speech Foundation Models Authors: Natarajan Balaji Shankar, Zilai Wang, Zihan Wang, Mohan Shi, Kaiyuan Zhang, Abeer Alwan Published: 2026-09-21 Categories: eess.AS, cs.CL arXiv: https://arxiv.org/abs/2609.23979v1 Attention encoder-decoder (AED) Speech Foundation Models achieve strong ASR performance but can generate acoustically unsupported text when inputs contain no speech, weak acoustic evidence, or unreliable transcription. We propose AURA: Activation-editing with Uncertainty-Routed Adaptation, an ultra-efficient representation-editing method that freezes the pretrained model and applies sparse scale-and-shift edits to decoder cross-attention heads. AURA dynamically routes edits using cross-attention uncertainty features that capture over-concentration, diffuse attention, and abrupt frame shifts. We evaluate AURA on four datasets spanning non-speech hallucination and speech grounding stressors, including imperfect-label child speech, imperfect-label adult speech, and disfluent speech. On non-speech audio, AURA reduces hallucination rate from 89.18% to 1.94% without prior hallucination-head identification. On imperfect-label corpora, AURA approaches LoRA WER while using roughly 500x fewer trainable parameters. Sensitivity analysis and qualitative cross-attention examples are consistent with AURA's uncertainty-routed editing behavior, supporting dynamic activation editing as a practical path for grounding AED speech models. --- ### LEAP-NBV: Lightweight Edge Active-Perception for Foundation-Model Next-Best-View Planning Authors: Boxun Hu, Jiawei Ge, Axel Krieger, Peng Wang, Tinoosh Mohsenin Published: 2026-09-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.23974v1 Foundation models are endowing autonomous systems with greater intelligence, enabling a more comprehensive understanding of the environment through visual perception. A representative example is Human Mesh Recovery (HMR), which provides useful estimates of a target's 3D pose and shape that can benefit tactical missions. However, the size and power demands of such models make them difficult to run on edge platforms and limit their real-time performance, undermining the requirements of tactical edge deployment - especially for active perception, where a mobile robot must plan its next-best view on-board and cannot offload computation under contested communications. We present LEAP-NBV, a lightweight active-perception framework that runs foundation-model-driven Next-Best-View (NBV) planning on-board an edge device. To this end, we distill a family of large HMR teachers, each into a compact 32M student, with an offline mesh objective, then quantize the vision encoder to FP16 and characterize its on-device accuracy and latency. Within an occlusion-aware active perception loop, we evaluate all configurations on the same held-out benchmark and deploy the end-to-end pipeline on an NVIDIA Jetson Xavier NX, reporting measured on-device latency and energy. Distillation recovers 6-7 mm of Procrustes-aligned mean per-vertex position error (PA-MPVPE) over the undistilled student on the test set. Selecting the edge-optimal compression model brings the HMR engine to ~12 ms at a small accuracy cost and runs the full closed loop at 3.6 FPS and 2.6 J per frame, achieving a 2.0x speedup and 3.0x lower energy than the uncompressed model while nearly matching downstream task quality. --- ### UniK: Universal Knowledge Perception for Digital and Physical AI Authors: Nirmit Desai, Kunal Sawarkar, Aditya Mahakali, Dongkon Lee, Kevin Park, Eric Song Published: 2026-09-21 Categories: cs.AI, cs.CV, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2609.23971v1 Two transformative classes of AI systems are reshaping how organizations operate: \textit{digital AI}, which reasons over enterprise knowledge to power chatbots and agent workflows; and \textit{physical AI}, which learns to control robots and autonomous systems from video, gameplay, and sensor telemetry. Both face the same foundational bottleneck: raw knowledge at scale, spanning heterogeneous modalities, locked in private corpora that existing AI infrastructure cannot access reliably or efficiently. We propose \textit{Universal Knowledge Perception (UniK)} as a common platform for both classes, covering the full knowledge lifecycle (ingestion, enrichment, indexing, retrieval, and continuous evaluation) across modalities from rich text and video to molecular data and sensor telemetry. We present UniK, built on Polymath Retrieval (multi-index fusion over automatically enriched indices) with no task-specific fine-tuning. Across five digital AI domains (medical literature, open-domain QA, chemistry, legal video proceedings, and government open data) UniK combined with an open-source 70-billion-parameter model consistently matches or outperforms frontier proprietary LLMs that are orders of magnitude larger: 76\% RAG accuracy on government data versus 47\% for GPT-5; 77.9\% on medical QA without fine-tuning; topping all open-source chemistry pipelines. We show that the same infrastructure directly addresses the data curation, indexing, and retrieval challenges facing physical AI world model training, where the knowledge problem is harder but structurally identical. --- ### Exponential Family Synthetic Controls Authors: Hector Rodriguez-Deniz, David M. Blei Published: 2026-09-21 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2609.23970v1 We develop exponential family synthetic controls (EFSC), a distributional version of synthetic controls for a panel of datasets. Each cell of the panel corresponds to a dataset drawn from an exponential family whose natural parameters factorize probabilistically across units and times. We estimate the latent factors using black-box variational inference. This replaces the usual weighted-average view of synthetic controls with a flexible probabilistic model that operates on full distributions. We propose causal estimands based on divergences between pre- and post-intervention distributions induced by the posterior of the natural parameters, together with distributional placebo tests to support causal inference and assess the significance of the estimated effects. We validate the proposed framework on synthetic and real data. Across a variety of exponential-family distributions, EFSC accurately recovers causal effects induced by exponential tilts, together with the corresponding divergences between treated and counterfactual distributions. The framework also captures effects induced by structural perturbations of the latent factors and by heavy-tailed noise contamination. Finally, we apply EFSC to study the expansion of Medicaid under the Affordable Care Act (ACA) and its impact on the distribution of health insurance coverage across U.S. states. Code is available at https://github.com/blei-lab/efsc. --- ### Rethinking Diffusion Segmentation: When Does It Rely on Its Noisy State, and Does Diffusion Matter? Authors: Hengzhuo Yang, Yuming Zeng, Yuling Yang Published: 2026-09-21 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.23967v1 Diffusion models are increasingly adapted from generation to conditional prediction, where a conditioning signal is combined with an evolving noisy representation of the target. In fully supervised segmentation, however, the conditioning image can already support direct target prediction, so endpoint performance alone establishes neither reliance on the added diffusion state nor a deterministic advantage over image-only prediction. For state reliance, we disrupt target-derived state content or correct image-state pairing during retraining of twelve published methods across three datasets, with ten matched seeds per setting. All 40 original-method comparisons whose evaluated-mask routes remained downstream of noised-quantity reconstruction exhibited state reliance, whereas all 30 comparisons with a segmentation-supervised bypass preserved reference performance. Rerouting five originally bypass-capable methods by forcing segmentation supervision through noise-to-mask reconstruction converted all 30 corresponding comparisons from preserved performance to state reliance. For deterministic utility, matched image-only counterparts achieved similar or better performance in 28 of 35 settings overall, including 16 of 20 whose native methods relied on both audited state properties. These results identify supervision path as a determinant of state reliance in the audited methods. Separately, matched image-only counterparts show that diffusion-specific computation often provides no deterministic endpoint advantage, including in methods that rely on the audited state properties. More generally, when conditioning already supports strong target prediction, diffusion-specific claims require additional evidence that the added state is used and that diffusion-specific computation improves the claimed capability beyond a matched condition-only counterpart. --- ### From Tables to Quantified Statements: Evaluating LLM Inference Generation through Executable Verification Authors: Mai Mohamed Eida, Gunjan Anand, Ayush Singh, Aleksandre Maskharashvili Published: 2026-09-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.23966v1 LLMs can generate fluent descriptions from tables, but their outputs may remain logically unsupported by the structured data. We introduce STAT-TO-TEXT, a controlled task in which LLMs generate quantified natural language inferences from statistical tables using quantified constructions such as all, some, no, and most. To evaluate these inferences, we use an LLM generated Python checker code which when executed verifies the corresponding truth conditions against the table. We compare four open-weight LLMs across model families and scales, evaluating faithfulness, logical accuracy, table coverage, and diversity. Our results show that model scale and family matter, with the largest model (GPT-OSS-120B) consistently producing the most faithful inferences without sacrificing greater table coverage and quantifier diversity, as opposed to smaller models. These findings are supported by human annotation, which shows that the automated checker closely aligns with human judgments. --- ### Colon3R: Cross-Domain 3D Reconstruction from Monocular Colonoscopic Video Authors: Zhihao Xing, Yingyu Wang, Liang Zhao, Shoudong Huang Published: 2026-09-21 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.23961v1 Monocular colonoscopic 3D reconstruction is important for surgical robotic colonoscopy, but remains challenging due to weak texture, specular reflections, limited view overlap, and non-rigid tissue motion. Conventional multi-view 3D reconstruction methods rely on stable correspondences and approximate rigidity, which are often violated in colonoscopy. Existing endoscopic methods often rely on domain-specific supervision, whereas there are not enough in-vivo labeled data available to adapt geometry foundation models to clinical colonoscopy. We present Colon3R, a cross-domain semi-supervised framework built on pretrained VGGT that transfers coupled camera, depth, and pointmap geometry from labeled phantom and simulated data to unlabeled in-vivo colonoscopy without requiring target-domain geometric annotations. Unlike source-only fine-tuning, which learns only from phantom and simulated data, Colon3R directly exploits unlabeled in-vivo video through teacher-derived cross-view supervision. Our proposed hierarchical quasi-rigid reliability selects reliable supervision at the sequence, directed-pair, and pixel levels, while source-preserving adaptation retains the learned coupled geometry during target-domain adaptation. Extensive experiments demonstrate that our method achieves superior overall performance over state-of-the-art approaches in depth, pointmap, and camera pose estimation. Qualitative comparisons on real in-vivo colonoscopy further show substantially more complete and geometrically consistent reconstructions than competing methods under clinical domain shift. The code will be public available after the paper is accepted. --- ### Open-Jev Judgments on CallScreenBench: Calibrated One-Pass Scam Screening with a Small Language Model Authors: Simiao Ren, Kidus Zewde, Xingyu Shen, Yuchen Zhou, Dennis Ng, Ankit Raj, Tommy Duong, Yuxin Zhang, Neo Tiangratanakul Published: 2026-09-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.23959v1 Screening a phone call for fraud needs a trustworthy probability after every caller turn, in milliseconds. Jev-style typed decisions promise exactly that: declared options go in, one calibrated probability per option comes out of a single forward pass, with no generated text. We test an open implementation of this readout, JevLite, on scam-call screening: Qwen3-4B is LoRA-tuned so that the temperature-scaled softmax over two answer-label logits is P(scam). On 41 held-out CallScreenBench scenarios (577 per-turn decisions) a three-seed ensemble reaches AUROC .974 with calibration error .052, non-inferior to an LLM judge (MiniMax-M3) at a pre-registered .02 margin, with no false alarms on legitimate calls, decisions 1.14 turns earlier under the same hang-up rule, and 64.5 ms per decision on one consumer GPU, 4.9x lower than the same backbone fine-tuned to generate its answer. The gain is in the readout and calibration, not accuracy: a fine-tuned ModernBERT encoder is not significantly worse, the recipe was selected with test-set exposure, and all callers are synthetic. We claim no architectural novelty; the contribution is the application and an evaluation reporting calibration, false alarms and decision timing alongside AUROC. --- ### Divergent strategies and convergent outcomes in autonomous materials discovery Authors: Jihan Kim Published: 2026-09-21 Categories: cs.AI, cond-mat.mtrl-sci arXiv: https://arxiv.org/abs/2609.23957v1 Scientific agents are mostly evaluated on whether they complete tasks or recover known results; we instead study variation across repeated open-ended campaigns. Sixteen separately initialized sessions of one model-harness configuration received a frozen database of 12,499 metal-organic frameworks, a methane-storage objective, a pinned protocol and a one-week budget. Strategies diverged into four approaches spanning 100--5,000 screened structures, and eight built 2,253 hypothetical structures. Yet the agents recovered the same materials frontier near 200 cm^3/cm^3, and an independent calculation of the database's porous region found its nine best structures all among their reports. Enforced checks on half the agents raised fresh-run reproduction from one of eight to eight of eight but could not detectably improve conclusion validity, because fifteen of sixteen agents selected the same audit-excluded entry, an incomplete structure whose missing anions created artificial pore volume. Replicated agents thus reveal both robust conclusions and common-mode errors from shared inputs. --- ### Some Dialects Are More Equal Than Others: Non-Prestigious Arabic Dialectal Bias in LLMs Authors: Mai Mohamed Eida, Ryan Dolan, Paul de Nijs, Jonathan Dunn Published: 2026-09-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.23955v1 Previous work on Egyptian Arabic in NLP has focused largely on the prestigious Cairene Egyptian Arabic (CEA) dialect, resulting in a lack of representation for the less prestigious Sa'idi Egyptian Arabic (SEA) dialect both in LLM and resource development. Does this lack of representation influence an LLM's view of the acceptability of SEA (upstream), and does an upstream bias against SEA lead to worse performance (downstream)? We investigate the upstream effect of SEA dialectal features on LLM preferences in a Targeted Syntactic Evaluation (TSE) task which reveals a significant bias against SEA across multiple LLMs. We then analyze the effect of these same features on downstream model performance on MMLU benchmarks and show that models experience a degradation in performance when presented with SEA. This work highlights the need for further exploration on how sub-dialectal variation impacts language technologies. --- ### Djinnlang: Higher-Level Programming by Unambiguous Specification with an LLM in the Compiler Authors: Simon Henniger, Stephen Chong, Nada Amin Published: 2026-09-21 Categories: cs.PL, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2609.23954v1 Programmers write formal specifications, and LLMs implement them, proving that each implementation matches its spec. Taken to its extreme, this makes specification languages the new programming languages. We argue that an unambiguity constraint is key: in addition to proving that its implementation satisfies the specification, the LLM must also prove that any other implementation satisfying it must produce the same outputs on the same inputs, i.e. that the relation formed by the constraints is deterministic. This leaves the LLM no leeway on program semantics: as with a conventional compiler, the generated code never needs to be read and can be regenerated from the spec at any time. Under this constraint and with a powerful LLM, the difference between a specification language and a programming language becomes essentially meaningless, and the LLM essentially becomes a part of the compiler toolchain. The arrangement doubles as a strong form of AI control: an untrusted model writes the code, yet its work is tightly checked by a verifier. To demonstrate that our LLM-in-the-compiler paradigm is feasible when supported by our unambiguity constraint, we present Djinnlang, a high-level specification language built for this future. A Djinnlang program consists only of specifications --- the programmer never writes executable code. In place of a traditional compiler, a symbolic translator lowers each spec to Dafny stubs and proof obligations, and a driver harness orchestrates an LLM that fills in implementations and proofs, all checked by the Dafny verifier. We evaluate our language and implementation on multiple examples and we show that it is self-hosting: an LLM can implement the Djinnlang translator from its specification and the reimplementation can verify itself. --- ### PrimeSeeker: Capability-Oriented Supervision for Deep Search Agents Authors: Linzhi Peng, Hanting Chen, Heng Chang, Ke Cheng, Bowen Du, Weifeng Lv Published: 2026-09-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.35816v1 Large language model search agents are often trained with synthetic questions whose difficulty is increased through larger evidence graphs, additional hops, and longer trajectories. These global properties, however, are only indirect proxies for the local retrieval capabilities required during search. To address this mismatch, we introduce latent anchor reasoning, which consists of resolving an unnamed retrieval anchor from descriptive specifications and transferring the recovered anchor into a subsequent information demand. This primitive retrieval unit decomposes deep search into chains of coupled operations and organizes question construction around anchor resolution and relation transfer, without prescribing a canonical search path. Based on this formulation, we propose PrimeSeeker, a capability-oriented framework that constructs web-grounded anchor structures and jointly derives a question and a reference evidence skeleton. The skeleton preserves supporting evidence from construction and guides expert generation through extractive highlights of current tool observations. These highlights are removed before supervised fine-tuning, while the skeleton is subsequently reused to audit reference-step coverage for reinforcement-learning rewards. We construct 9,221 expert trajectories, training a 30B search agent. Across five deep-search benchmarks, PrimeSeeker achieves strong performance, while reference-step optimization further improves the supervised policy. The resulting trajectories exhibit low retrieval redundancy, and fixed-budget evaluation shows strong solution coverage with substantially fewer tool calls than long-horizon systems. --- ### How to Run Statistics over LLM Judges and Trust the Results: Calibrated Inference for Small-Sample AI Evaluation with evalstats Authors: Ian Arawjo Published: 2026-09-20 Categories: cs.CL, cs.AI, cs.HC, stat.ME arXiv: https://arxiv.org/abs/2609.35815v1 Researchers across academia increasingly base significance claims on LLM judge scores and small-sample AI evaluations. Yet without well-calibrated confidence intervals (CIs), hypothesis tests, and judge-bias corrections, such claims are unreliable. We address these issues in several contributions. First, we find that running statistics over raw LLM judge scores leads to inflated false positives: counterintuitively, for many inter-rater agreement metrics, false positive risk peaks at "almost perfect" human-LLM agreement. To help researchers understand how to run statistics over LLM judges responsibly, we present guidance and tooling for the statistical analysis of mixed human-AI judge designs, and implement nine hypothesis tests via prediction-powered inference (PPI), including the first known PPI corrections for four rank-based tests (Wilcoxon signed-rank, Mann-Whitney U, and omnibus variants). To keep PPI++ stable with small human-labeled calibration sets, we introduce bootstrap-adaptive power tuning, which shrinks the estimated weight toward a target estimated from the labeled data, and accounts for that weight's own sampling variance. Second, through Monte Carlo simulations, we derive recommendations for what CI, p-value, and FWER correction methods to use for small-sample AI evaluations (N<100), and warn researchers against bootstrap CIs. We package these recommendations into evalstats, an open-source Python package that selects calibrated methods automatically, and demonstrate it in three scenarios, including one where a real LLM judge validated at "substantial agreement" would have led a researcher to publish a spurious finding. evalstats is publicly available at https://github.com/ianarawjo/evalstats. --- ### Constructing Challenging Browser-Use Tasks by Controlled Environment Interventions Authors: Xunjian Yin, Tianchen Guan, Jinao Wang, Weili Cao, Daisy Xinlei Lin, Royce Cheng-Yue, Keagan Long, Kyle Wong, Bhuwan Dhingra, Xiangjun Wang, Shuyan Zhou Published: 2026-09-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.35814v1 As browser-use agents improve, benchmarks keep pace by collecting new tasks, websites, and applications, often making tasks longer or more novel. This makes difficulty expensive to refresh and difficult to control: when many aspects change at once, it is unclear what actually makes a task challenging. We instead construct challenging instances from tasks agents already solve, turning difficulty into a programmable property of the environment. BreakingWeb pairs every base task with an intervention condition that preserves the user instruction, latent target, and backend success criterion while changing the environment at different web stack layers. Each intervention is deterministic, detectable, and recoverable, and is annotated with the cognitive primitive it primarily loads. The benchmark contains 519 clean/intervention task pairs across seven self-hosted websites and 29 intervention families, all graded against outcomes. We evaluate six strong browser-use agents, three GUI-only agents that see only screenshots, and humans. The construction is effective: interventions cut agent pass rate by 22.9% on average and overturn nearly half of the tasks each agent solves cleanly, whereas humans lose 10.0% on a first attempt and 5.7% after one familiarisation attempt. The dominant failure is belief failure: 75% of the six agents' failures end with a declared success although the required change never happened. Our code, data and environment are publicly available at www.breakingweb.app. --- ### Local Predictability and Collective Fidelity in LLM-Agent Societies Authors: Igor Itkin Published: 2026-09-20 Categories: physics.soc-ph, cs.AI, cs.CL, cs.MA arXiv: https://arxiv.org/abs/2609.35813v1 Compact surrogates could reduce the cost of simulating large language model societies, but must reproduce collective behavior. We compare individual predictions and collective forecasts using 9,455 published trajectories and new experiments on opinion dynamics. Neighbor information improves individual prediction in all 16 public-data settings and pooled collective forecasts on held-out questions, although collective gains depend on transfer conditions. Tests on 24 new statements do not confirm earlier contrasting history effects in forecasts from the initial state. Qwen benefits from history after three observed rounds. These findings motivate direct collective validation, explicit limits on available observations, and comparisons with simple baselines. --- ### Automated Evaluation of Multi-Turn Dialogues in In-Car Conversational Assistants Authors: Vaishnav Negi, Lev Sorokin, Soroosh Tayebi Arasteh, Andrea Stocco Published: 2026-09-20 Categories: cs.CL, cs.SE arXiv: https://arxiv.org/abs/2609.35812v1 In-car conversational assistants (ICAs) are increasingly integrated into vehicles to support route planning, vehicle control, and information access. Ensuring their reliability is challenging due to multi-turn interactions, the absence of explicit ground truth, and strict safety constraints. Existing evaluation techniques fall short, as they target single-turn settings and fail to capture constraint handling, context retention, and safety-critical behavior across turns. We propose an automated framework for testing the multi-turn conversational capabilities of ICAs. The system is treated as a black box and evaluated via closed-loop simulation with a strategy-guided user simulator, an adversarial strategy manager, and a two-tier LLM judge assessing turn-level failures and conversation-level quality. We evaluate the approach on an industrial ICA with six LLM backends and twelve human annotators. The automated judge shows substantial agreement with humans, and strategy guidance uncovers 2.96 times more unique failure types per conversation and more than doubles the number of unique failing conversations compared to unguided simulation. --- ### Lookahead-R: Budget-Aware Tool Retrieval via Execution-Centric Planning Authors: Zongze Wu, Yani Guo, Runnan Li Published: 2026-09-20 Categories: cs.CL, cs.AI, cs.SE arXiv: https://arxiv.org/abs/2609.35811v1 Tool retrieval is a critical bottleneck for LLM-based agents operating over large, heterogeneous API ecosystems. Existing approaches face an inherent trade-off: semantic retrievers are fast but suffer from the semantic-functional gap, while execution-based validation improves precision at the cost of prohibitive latency. We propose Lookahead-R, a planning-based framework that reformulates tool retrieval as a resource-constrained sequential decision-making problem. At its core, Lookahead-R introduces a lightweight execution-aware surrogate world model that jointly predicts tool execution success, latency cost, and semantic utility---without invoking real APIs. This world model drives a cost-sensitive, uncertainty-guided Monte Carlo Tree Search that navigates the tool space under strict budget constraints. Evaluated on the large-scale ToolBench benchmark, Lookahead-R achieves a superior accuracy-efficiency trade-off across all test scenarios. On the most challenging I3 split, it attains an NDCG@5 of 91.40\%, outperforming the state-of-the-art ToolGen (90.16\%) by 1.24\%. Ablation studies confirm that explicit latency modeling is the key discriminative signal for identifying high-quality tools under resource constraints. --- ### TRACE: Deployable Tree-Relational Structure Enhancement for Oncology LLMs Authors: Jizheng Lai, Yingyun Li, Ying Qin, Haiyang Qian Published: 2026-09-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.35810v1 Large language models are increasingly used in oncology applications, but their predictions are often weakly grounded in explicit medical structure. We present TRACE, a deployable tree-relational enhancement framework for oncology LLMs. TRACE separates expensive offline structure learning from lightweight online inference: oncology concepts and relations are organized into an updatable tree-relational structure, refined using LM-loss-derived evidence, and retrieved at inference time as compact prompt evidence. This design supports task-adaptive evidence selection without requiring supervised labels in the zero-shot setting. Across ten oncology classification tasks and one MedQuAD CancerGov QA benchmark, TRACE improves both label-free evaluation and supervised fine-tuning. Additional analyses show that TRACE improves over vanilla RAG and generic GraphRAG, remains useful under leakage-controlled METABRIC inputs, and produces interpretable evidence paths aligned with clinical reasoning. These results suggest that explicit, updatable medical structure is a practical path toward more accurate and auditable oncology LLM deployment. --- ### Can Multimodal Large Language Models Generate and Detect Multimodal Social Media Fake News? Authors: Jiyao Yang, Yang Liu, Zhenyue Qin, Qingyu Chen, Xiuzhen Zhang Published: 2026-09-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.35809v1 The rapid advancement of generative AI raises concerns about the misuse of Multimodal LLMs (MLLMs) for large-scale disinformation campaigns on social media. Despite existing research on textual disinformation, a fundamental question remains unanswered: can MLLMs be exploited to fabricate realistic multimodal fake news, and can they reliably detect it? We introduce a multi-agent framework in which a story agent, an image agent, and a critic agent collaborate to produce fake social media posts that plausibly counter true news. We apply the framework to generate over 9,000 paired multimodal news posts across science, health, and entertainment domains, and benchmark 16 open- and closed-source MLLMs for automated detection. We find that most models fall substantially short of human-level accuracy and fail critically on identifying image authenticity. Our research provides a foundation for developing robust defenses against social media fake news. Code and data are available at https: //github.com/xiuzhenzhang/Multimodal. --- ### When Successful Memories Mislead Embodied Agents:Memory Adaption For Task-Conditioned Execution Authors: Quanquan Li, Hongbo Zhang, Yihe Chi, Liuyang Song, Jingyu Li, Yuxiang Huang, Hongzhen Zhang, Guitao Cao Published: 2026-09-20 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.35808v1 Experience reuse can reduce repeated exploration in embodied agents, but a trajectory that succeeded previously may be unsuitable for the current execution context. Existing memory systems pri marily optimize construction and retrieval; semantic relevance and historical success therefore remain insufficient when retrieved ex perience contains incompatible actions or an inappropriate level of structure. We introduce Memory Adaptation for Task-Conditioned Execution (MATE), a deterministic post-retrieval procedure that converts trajectories into execution-oriented memory. MATE re moves obsolete control context, extracts condition-action-effect transitions, applies verified action normalization, selects a task dependent representation, and serializes the result under a fixed budget without additional LLM inference. On 134 ALFWorld tasks, MATE achieves task success rates of 81.3% and 93.3% with Qwen2.5-14B and 72B while using approximately one-tenth of the tokens required by raw trajectories. Controlled comparisons show that verified action normalization is the principal mechanism by which MATE restores the utility of retrieved experience, support ing memory adaptation as a distinct stage between retrieval and embodied execution. --- ### Cross-Dataset Generalization of Bangladeshi Rice Leaf Disease Classifiers: Benchmark, Diagnosis, and Mitigation Authors: Anindya Paul Published: 2026-09-20 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.31709v1 Cross-dataset transfer in rice leaf disease classification remains a significant challenge, with models trained on one image collection performing substantially worse when deployed on another. We conduct a systematic benchmark across three Bangladeshi rice leaf disease datasets (5,419 images, 6 transfer pairs, 3 CNN backbones, 3 random seeds) to characterize and diagnose this failure. Strong augmentation recovers a mean cross-dataset macro-F1 improvement of +0.070 (Wilcoxon p < 0.001, 15 of 18 transfer pairs positive). Removing non-leaf image content via segmentation shows directional benefit (mean +0.066, p = 0.062, n = 36 paired observations) that is consistent across two independent segmentation methods but does not reach conventional significance. A self-supervised ViT control (DINOv2 linear probe) exhibits equivalent cross-dataset collapse to CNNs, ruling out architecture inductive bias as the primary driver and pointing to acquisition-condition shift. Adaptive batch normalization uniformly harms transfer performance, with harm magnitude correlating with source-target label-prior divergence and model depth (Spearman rho = 0.621, p = 0.009). Grad-CAM attribution analysis on 12 sampled predictions does not distinguish correct from incorrect cross-domain predictions (p = 0.462), indicating that common attribution proxies are insufficient for diagnosing shift at practical sample sizes. We document all frozen results, prespecified analysis criteria, and reproducibility artifacts in a public repository with SHA-256 integrity verification. This work establishes a rigorous empirical baseline for understanding cross-dataset generalization in agricultural computer vision and identifies both effective (augmentation) and ineffective (AdaBN) adaptation strategies. --- ### Can't Find Waldo: Evaluating VLMs' Sensitivity to Image Resolution and Detail Level Authors: Alexandra Schild, Gerard de Melo Published: 2026-09-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.31706v1 Visual Language Models (VLMs) have achieved remarkable success across diverse tasks, yet they struggle with high-resolution inputs where critical information resides in small regions or detailed, cluttered scenes. While several approaches address this limitation, a systematic understanding of why models fail at high resolutions is lacking. We introduce a controlled evaluation framework that disentangles resolution-related performance degradation from task difficulty through semantics-preserving transformations. We propose two simple metrics: Area Under the Scaling Curve (AUSC), which quantifies scaling robustness independent of baseline accuracy, and Prediction Variance Score (PVS), which measures resolution-induced prediction instability. Through comprehensive experiments across 5 model families and 5 benchmarks, we identify three primary failure modes: (1) information loss from downsampling at vision token limits, (2) tokenization artifacts from patch boundary shifts and positional encoding fragility under non-standard aspect ratios, and (3) attention dilution as token counts increase. Our analysis reveals that even state-of-the-art models suffer from performance drops when processing high-resolution images, with degradation patterns varying systematically by architectural family. We provide actionable insights for model architecture design and data augmentation strategies to mitigate these limitations. --- ### MDL-Calibrated Significance-Gain Pair Encoding: Replication-Aware Automatic Stopping for Subword Tokenization Authors: Azam Nouri Published: 2026-09-20 Categories: cs.CV, cs.CL arXiv: https://arxiv.org/abs/2609.31705v1 Byte-Pair Encoding (BPE) constructs subword vocabularies through greedy pair merging, but conventional BPE requires the number of merges or target vocabulary size to be specified externally. Significance-Gain Pair Encoding (SG-BPE) replaces frequency-only selection with a statistical criterion based on how strongly an observed pair exceeds its expected co-occurrence under an independence model. This paper introduces MDL-Calibrated Significance-Gain Pair Encoding (MDL-SG), a three-stage procedure separating discovery, replication, and utility. Candidate pairs are ranked by Significance-Gain on a discovery partition, tested for replication on a separate partition using an exact one-sided hypergeometric test with per-iteration Benjamini-Hochberg correction, and then evaluated on a utility partition using a Minimum Description Length (MDL) criterion. Merging stops automatically when no replicated candidate yields positive held-out MDL gain. On WikiText-103, MDL-SG stops at 209, 433, and 847 merges for 120K, 250K, and 500K-character tokenizer-training samples, respectively. At 500K characters, it selects a stored vocabulary of 1,017 tokens without prescribing the vocabulary size in advance. In a compute-matched TinyGPT experiment with identical 2,024,448-parameter models and 500 optimizer updates per language model, MDL-SG achieves validation/test BPC of 3.2612/3.2436, compared with test BPC of 3.2894 for SG-BPE and 3.3493 for frequency BPE. Frequency BPE achieves stronger raw compression, while MDL-SG achieves lower BPC, showing that compression-oriented merge selection and language-model utility need not coincide. --- ### Does DCGAN-Based Synthetic Augmentation Improve Brain Tumor MRI Classification? An Empirical Study Authors: Irhum Jawad Khan, Talha bin Aslam Published: 2026-09-20 Categories: eess.IV, cs.CV arXiv: https://arxiv.org/abs/2609.28508v1 Generative adversarial networks (GANs) are increasingly used to augment medical imaging datasets, but synthetic images do not necessarily provide downstream classification benefits. This study investigates whether class-specific Deep Convolutional Generative Adversarial Network (DCGAN) augmentation improves brain tumor classification when the classifier and evaluation set are held constant. Experiments were conducted on 7,200 brain magnetic resonance imaging (MRI) scans across four classes: glioma, meningioma, pituitary tumor, and no tumor. For each class, 1,400 real images were used for training and 400 were reserved for testing. A baseline Swin Transformer classifier was trained using only the real training images and compared with a second model trained using the same real images augmented with 500 DCGAN-generated images per class. Both conditions were evaluated on the identical held-out test set. The two models achieved the same overall accuracy of 96%, while macro F1 remained effectively unchanged and ROC-AUC decreased slightly from 0.987 to 0.982 after augmentation. Class-level analysis showed small redistributions in errors rather than a consistent performance gain. FID values between real and synthetic images ranged from 209.15 to 314.27, indicating substantial distributional differences under the adopted evaluation setup. These results suggest that synthetic augmentation should not be assumed to improve medical image classification and should instead be evaluated for both distributional fidelity and downstream task utility. --- ### Bridging LLM Serving and CXL-SSDs with Chunk-Aware KV Cache Management Authors: Hyunsun Chung, Taewan Noh, Minji Kim, Joo-Young Hwang, Hong-Yeon Kim, Youngjae Kim Published: 2026-09-20 Categories: cs.AR, cs.AI, cs.ET arXiv: https://arxiv.org/abs/2609.26828v1 NAND-backed storage offers the capacity needed to scale LLM prefix caching, but its block I/O path incurs CPU cache contention and host-DRAM staging in addition to NAND latency. Our characterization shows that these interface costs persist even with DRAM as the storage medium, motivating CXL-SSDs for byte-addressable access to NAND-backed capacity. Surprisingly, however, a stock CXL-SSD remains about 3$\times$ slower than local DRAM and no faster than an NVMe SSD, while generic prefetching provides little benefit. We present LM-CXD, a CXL-SSD specialized for LLM prefix caching. LM-CXD bridges the semantic gap between the serving engine, which knows which KV chunks will be consumed, and the device, which controls their placement and movement. It makes KV chunks device-visible I/O units, exposes NAND-to-DRAM progress to the serving engine, and uses device DRAM as a GPU-accessible buffer. LM-CXD further coordinates request scheduling with windowed prefetching and pipelines layerwise KV movement with GPU computation to hide NAND latency under limited device DRAM. Across five LLM models, LM-CXD reduces average TTFT over a stock CXL-SSD by up to 2.6$\times$ with compute asynchronous prefetching and 4.03$\times$ with layerwise prefetching, achieving TTFT within 1.5$\times$ of local DRAM on average. --- ### What Makes a Terminal-Bench Task Hard? Separating Genuine Hardness from Fake-Hardness on an Adjudicated Agentic Corpus Authors: Edward Lue Chee Lip, Boden Moraski, Tim Knappe, Lang Xiong, Sarvesh Gharat, Antonio Mari, Ivan Bercovich Published: 2026-09-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.26826v1 Frontier benchmarks need tasks that current models cannot solve. But a task that no model solves is not automatically a hard task. The same zero pass rate can come from a real capability gap, but it can also come from missing context, a broken reference solution, infrastructure failure, or a verifier that can be bypassed. In this paper, we study this issue using a frozen Terminal-Bench 3 / Frontier-Bench 0.1 production record with 1,081 pull requests, 639 scored tasks, 28,801 trials, and $105,933 in logged agent spend. We ask what an all-fail task actually certifies. For the 125 tasks with no honest pass, we combine task artifacts, reference-solution runs, empty-solution controls, adversarial trials, trajectories, telemetry, and review records, and apply an ordered validity screen. Only 78 of the 125 tasks survive as certified-unsolved candidates. The remaining tasks include 14 with broken oracles, 8 dominated by infrastructure failures, 4 that are only passable through verifier bypasses, and 21 whose solvability is not certified by the available evidence. Thus, lack of saturation and genuine difficulty are not the same thing. The certified-unsolved label is also narrow: it means that the authored route passed, infrastructure did not dominate, no strict bypass was observed, and all evaluated agents failed. It does not prove intrinsic hardness, verifier completeness, or failure at the intended capability. We further analyze rejected submissions and passing tasks to show that pass rate alone cannot explain why a task is difficult. Overall, our results suggest that frontier benchmarks should report the evidence behind their all-fail tasks before using them as capability claims. --- ### Text Scores Can Miss Waveform Use: A Qwen2-Audio Quantization Case Study Authors: Mengzhe Geng, Jinxi Jin, Junhao Xu Published: 2026-09-20 Categories: cs.SD, cs.CL, eess.AS arXiv: https://arxiv.org/abs/2609.26823v1 Post-training quantization of speech language models is often summarized with text-output scores and nominal bit widths. Those numbers alone do not establish behavior that depends on information missing from a transcript, or efficiency for a particular runtime. We introduce an evaluation protocol that separately tests lexical output, a transcript-insufficient endpoint, and a measured packed implementation. In a Qwen2-Audio case study, a translation-selected 6-bit allocation improves chrF by 2.36 on a frozen English-to-German replay, with paired 95% bootstrap interval (1.04, 3.62), but loses 3.91 percentage points on speaker-disjoint emotion recognition. At the same 6-bit budget, the uniform structural control reaches higher emotion accuracy than the selected allocation, and the front-layer control is also higher by point estimate on the same frozen set. At 7 bits, chrF improves by 3.28 with interval (2.08, 4.59), the emotion interval against FP16 includes zero, and a same-budget front-layer control still exceeds the selected allocation. A separate matched-budget 4.08-bit study finds roughly 10-point emotion deficits for every tested low-bit allocation and no selected-allocation advantage over frozen controls. Finally, a dequantized average-6-bit simulation retains the FP16 peak memory. This case study identifies a precision-dependent mismatch between lexical output, waveform-dependent behavior, and nominal precision. It does not establish a general failure of low-bit speech models or a deployment benefit for the selected allocation. --- ### HARN: Hierarchical Associative Resonance Network for Event-Driven Multi-Timeframe Forecasting Authors: Nabeel Ahmad Saidd Published: 2026-09-20 Categories: cs.LG, q-fin.CP, q-fin.GN arXiv: https://arxiv.org/abs/2609.26822v1 Financial time series evolve across multiple temporal resolutions, challenging forecasting systems to incorporate newly available information without repeatedly recomputing unchanged representations. We introduce HARN, a Hierarchical Associative Resonance Network for event-driven multi-timeframe forecasting. HARN maintains persistent representations across temporal levels and updates each level only when its corresponding completed bar becomes available. The architecture combines causal multi-scale temporal encoding, gated associative memory, cross-level resonance, and hierarchical evidence aggregation, with forecasting performed in basis-point space and reconstructed to the original price scale. We evaluate HARN on four assets spanning equity, foreign exchange, and commodity markets using multiple random seeds and component ablations. HARN achieves competitive reconstructed-price forecasting errors against single-timeframe PatchTST and TimeXer baselines, while ablations reveal the effects of removing individual components across assets and timeframes. A code-level audit further examines consistency between the implementation and the defined event-driven causal protocol. The results position HARN as a persistent multi-timeframe forecasting framework rather than evidence of universal predictive superiority. --- ### Signal2Symbol: Neuro-Symbolic Temporal Reasoning for Explainable Physiological Time-Series Anomaly Detection Authors: Naser Mansour, Sidahmed Benabderrahmane, Ameer Rahwan Published: 2026-09-20 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.26820v1 Physiological time series such as electrocardiograms (ECG) and electroencephalograms (EEG) exhibit complex temporal structure, substantial acquisition variability, and a strong need for transparent decision-making. Although deep models can achieve high detection performance, they often provide limited insight into why a segment is anomalous, how local anomalies relate over time, and whether a detection belongs to a broader recurring pattern. We propose Signal2Symbol, a neuro-symbolic framework for explainable biosignal anomaly detection. The method first converts ECG/EEG signals into symbolic sequences using either a learned VQ-VAE (Vector Quantized Variational Autoencoder) codebook or a SAX (Symbolic Aggregate approXimation) baseline. It then constructs bigram enriched token-window transactions and scores anomalies through rare itemset evidence derived from minimal rare itemset mining. Detected anomalous windows are merged into intervals and related using Allen interval algebra, enabling composite temporal explanations such as escalation chains, artifact overlap, and cross-channel synchrony. Finally, we introduce a rare temporal concept lattice based on Formal Concept Analysis (FCA), which groups anomalous intervals by shared rare symbolic evidence, Allen temporal relations, channel context, and robustness attributes. The resulting Galois lattice compresses many local detections into interpretable families of temporal-symbolic anomalies. We evaluate on three public benchmarks: MIT-BIH Arrhythmia (beat-level ECG), PTB-XL (record-level ECG), and the Bonn EEG dataset (segment-level EEG). We stress-test robustness under additive noise and baseline-wander perturbations. The results highlight the value of neuro-symbolic tokenization for temporal anomaly analysis and show that Allen/FCA reasoning provides compact, interpretable summaries of local detections. --- ### MAC-RRG: Iterative Multi-Agent Collaboration for X-ray Radiology Report Generation Authors: Futian Wang, Yuhan Qiao, Xiao Wang, Dan Xu, Yuehang Li, Zhixiang Guo, Yaowei Wang, Jin Tang Published: 2026-09-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.26124v1 Despite the remarkable progress of LLM-based and knowledge graph-augmented Radiology Report Generation (RRG) methods, existing techniques still suffer from inherent defects. Conventional LLM-only models lack structured medical prior knowledge, resulting in frequent medical hallucinations and low diagnostic interpretability. Current knowledge graph-enhanced schemes adopt static one-round knowledge fusion with single-source knowledge, incapable of dynamic knowledge updating according to generation feedback. This paper proposes a novel Multi-Agent Collaborative iterative framework for X-ray Radiology Report Generation, termed MAC-RRG. Inspired by multi-agent technology, our framework constructs a closed-loop optimization paradigm based on task decoupling and collaborative reasoning. Specifically, the framework first generates a preliminary radiology report from input X-ray images via a vision encoder and a basic LLM. Subsequently, a multimodal knowledge graph (MM-KG) agent mines structured disease correlation and anatomical knowledge from medical knowledge graphs, while an auxiliary knowledge agent extracts unstructured domain knowledge from public medical databases. The multi-source knowledge acquired by dual agents is fused and embedded to guide the LLM in iteratively refining the initial report. Extensive quantitative and qualitative experiments on mainstream X-ray RRG datasets, including IU X-ray, MIMIC, and CheXpert Plus, fully verify the superiority of our proposed method. The source code and pre-trained models have been released on https://github.com/Event-AHU/Medical_Image_Analysis --- ### The Probabilistic Structure of Large Language Models Authors: Adnan Aboulalaâ Published: 2026-09-20 Categories: cs.LG, cs.AI, math.PR, stat.ML arXiv: https://arxiv.org/abs/2609.25134v1 This paper presents a probabilistic perspective on large language models (LLMs), developed with the aim of bringing together, in a single self-contained account, tools that are usually treated separately across the literature. LLMs are described through probability measures on the set of sequences of tokens, specified via their autoregressive conditional distributions. Training is formulated as a maximum-likelihood estimation problem, addressed by stochastic gradient methods, while text generation is viewed as the sequential simulation of the resulting stochastic process. The role of the asymmetry of the Kullback--Leibler divergence in text generation is examined in relation with characteristic phenomena such as hallucination and the distinction between statistical plausibility and truth. As a complementary illustration of the same viewpoint, we also discuss diffusion models, built around the score function, which cast generation not as sequential token prediction but as the simulation of a reverse-time stochastic process transforming noise into data both in discrete and continuous time. --- ### Entropy Can Flow, or It Can Guide. Be Entropy. LEDFlow: Introducing Entropy-guided Generation Order into Uniform Discrete Flow Authors: Tung Sum Thomas Kwok, Yidong Ouyang, Yingjia Wan, Ying Nian Wu, Zhijiang Guo, Oscar Leong Published: 2026-09-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.25131v2 Uniform discrete flow permits repeated updates at every generation position. While continued revision supports correction of wrong tokens, it also exposes correct intermediate predictions to later errors. An experiment on Sudoku puzzles shows that 9.4% of generated cells are correct at an intermediate step but incorrect in the final output. We introduce generation order into uniform discrete flow through selective absorption, which fixes chosen predictions while preserving the uniform flow velocity at active positions. To prioritize reliable predictions for absorption, we propose Low-Entropy Discrete Flow (LEDFlow), a training-free sampler that adaptively orders absorption by local entropy. By decomposing absorption error into joint dependence and conditional prediction terms, we show that, under entropy-error regularity, selecting the lowest-entropy positions under a fixed absorption count minimizes an upper bound on the conditional term. We further analyze sensitivity of global lookahead, whose worst-case decision-error bound grows with lookahead window under an imperfect denoiser. Across reasoning benchmarks, LEDFlow attains 0.845 Nikoli Sudoku solve accuracy, with largest gains on strongly constrained tasks. On a text-to-image generation benchmark it attains the best overall score among decode-time samplers, and on multimodal understanding it improves over the default sampler on all six benchmarks, at an inference cost comparable to standard flow sampling. --- ### Impact Is Not Invalidation: Ask About the Claim, Not the Diff Authors: Atul Anand Published: 2026-09-20 Categories: cs.CL, cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.25130v1 Memory systems for coding agents must decide, when a repository changes, which of their stored claims have become false. Content anchoring invalidates a claim whenever the artifact it came from changes, which fires constantly. Semantic-equivalence classification asks whether a diff preserves behavior, a question about the diff rather than about any stored claim. We show the second signal fails for a reason unrelated to model capability: asked whether a commit preserves behavior, five models spanning a 40x price range fire on 59-72% of real commits and reach precisions of only 0.291 to 0.329 against a 0.25 base rate. Asked instead whether one specific claim still holds, the same models on the same diffs reach 0.705 to 0.974. A control that hands the behavior-preservation judge the claim text, changing only the question, moves precision by 0.010 and 0.016; changing the question moves it by 0.49 and 0.65. We also compare against pytest-testmon, a deployed regression-test selector with coverage-derived dependency data: it reaches 0.868 recall at 0.415 precision, so near-complete knowledge of what a change can reach does not identify what it falsifies. Ground truth is execution, not annotation: a claim is a test function passing at commit t, and it has flipped if that same assertion text fails at t+1. Building this required an observation we did not find in prior work. On a CI-gated mainline a commit that leaves a pre-existing test failing cannot merge, so the naive construction has an empty positive class by design. We report 10,369 claims with 184 execution-verified flips mined from 23 Python libraries, splits held out by repository, a post-knowledge-cutoff split, a shuffled-diff null, a paraphrase control, and a leave-one-repository-out analysis over 17 repositories. --- ### WILSON - a pathology foundation model framework for patient-level analysis and diagnostic text generation Authors: Saghir Alfasly, Wataru Uegami, Sobhan Hemati, Wenchao Han, Xiaojia Tang, Kevin Thompson, Daniel Stone, Ghazal Alabtah, Saba Yasir, Michael R. Lucas, Eric W. Klee, Cheryl L. Willman et al. Published: 2026-09-20 Categories: q-bio.QM, cs.AI, cs.CV, cs.LG, eess.IV arXiv: https://arxiv.org/abs/2609.25123v1 Pathologists integrate morphology across magnifications and across the slides of a patient case, whereas pathology foundation models encode thousands of tiles from single slides and aggregate their features. Here we present WILSON, a vision--language foundation model that represents whole-slide images and multi-slide cases as single multi-magnification composite images, trained on approximately 189k slides from Mayo Clinic spanning 42 organs and 829 diagnostic entities using pathology reports as supervision. Without task-specific training, WILSON exceeded a dedicated case-level model on all internal cohorts (macro-F1 0.52 versus 0.38) and matched slide-level models up to 9.4 times larger at 272- to 2,155-fold lower compute. End-to-end fine-tuning on 508 triple-negative breast cancer cases improved histologic subtyping and stromal tumor-infiltrating lymphocyte grading by 0.16 and 0.11 macro-F1. WILSON retrieved matching diagnostic text at 75.6% recall@1 (PRISM, 58.1%) and generated captions closer to report-derived references than PRISM and PRISM2 on the internal cohort and on most external comparisons. Composite images thus offer a compact, clinically aligned computational unit for pathology. --- ### Rachel: A general-purpose language model directs and revises retrosynthetic routes Authors: Qisheng Li, Shunchao Jiang, Chen Qi, Xin Su, Da Han, Guangyong Chen Published: 2026-09-20 Categories: physics.chem-ph, cs.AI arXiv: https://arxiv.org/abs/2609.25118v1 Retrosynthetic planning advances through decisions that reshape the remaining chemical problem: a locally plausible disconnection can leave precursors whose chemoselectivity constraints complicate the rest of the route. Existing planners often channel model proposals through search or template procedures, leaving open whether a general-purpose large language model (LLM) can itself sustain and revise route strategy. We developed Rachel, a stateful environment that executes and checks LLM-directed chemistry but prescribes neither a search policy nor a stopping rule. Without supplied reference routes or route-level solutions, GPT-5.5 achieved strict closure for 111 of 120 PaRoutes120 targets and 24 of 25 targets in the separate RF25 difficult-target cohort. RF25 was drawn largely from studies published after GPT-5.5's reported knowledge cutoff. Closure required complete routes and independent source resolution of every terminal precursor after planning. On a shared PaRoutes subset, forward-model support exceeded that of most comparator methods, and Rachel received the highest mean overall route score from both method-blinded LLM evaluators. Recorded trajectories showed continued model-proposed chemistry, with revised strategies carried into subsequent steps. Replacing LLM route decisions with fixed policies reduced strict closure to 6-15/120 despite continued local chemical execution; restricting planning support also reduced closure in RF25. Within Rachel, a general-purpose LLM coordinated successive chemical choices and revised its strategy as earlier decisions reshaped the remaining problems. --- ### You've Seen Enough: Quality-Constrained Image Coding for Machines Authors: Khoa Pham-Dinh, Sanaz Nami, Hamed Rezazadegan Tavakoli, Moncef Gabbouj, Farhad Pakdaman Published: 2026-09-20 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.25108v2 Visual data is increasingly consumed by machine-vision systems rather than by human observers. Image Coding for Machines (ICM) compresses images assuming the main observer is a computer vision application and that the human observer needs to inspect or validate the decisions. Inspired by just-noticeable distortion, we cap human-observed quality at a desired level and devote the remaining bits to machine performance. Specifically, joint compression-segmentation training is recast as a constrained optimization problem in which the codec must meet a predefined acceptable target visual quality while a task term consumes the remaining coding capacity. This paper proposes two variants of a penalty function that guides the quality toward the target: an absolute function and a bilinear function, the latter applying a steeper slope once the target visual quality is exceeded. Experimental results show that, under the quality constraint, the proposed method achieves BD-rates of $-22.82\%$ and $-29.81\%$ relative to an unconstrained joint rate--distortion--task optimization and a simple rate--distortion baseline, respectively, showcasing bitrate reduction with the same task performance. This is achieved while the codec also meets the target visual quality with a reasonable error and without adding any complexity overhead. --- ### Agents That Edit Documents: Measuring Agentic PDF Forgery Against a Non-Agentic Control Authors: Simiao Ren, Ankit Raj, Tommy Duong, Yuxin Zhang, Dennis Ng, Xingyu Shen, Kidus Zewde, Yuchen Zhou, Neo Tiangratanakul Published: 2026-09-20 Categories: cs.AI, cs.CR arXiv: https://arxiv.org/abs/2609.23953v1 AI agents that carry a multi-step computer task through on their own became ordinary tools in the past year, and the same autonomy is available to anyone whose task is harmful. We ask what that means for a relying party -- an insurer, a lender, an auditor -- whose evidence is a filed PDF. AgentForge-Bench measures how reliably an off-the-shelf coding agent, driving one of seven open-weight models with a shell and the stock Python PDF stack, alters one dollar amount, date or address in a real filed financial document from a single sentence of intent, graded by rules rather than by a model. Across 1,750 cells, 1,419 (81.1%) satisfy the verifier, and 808 (46.2%) also survive every stricter filter: visible, localized, typeface-matched, original value gone document-wide. A deterministic script with no model in it solves 98 of the 125 documents; the agents solve 124, and none the script solves alone. Agents misreport 41% of their wrong edits as done, no model refused, and the cheapest verified forgery costs 2.4 cents. The raw rate overstates the threat by about a factor of two; the strict rate is still large. --- ### HaikuS2S: A Cascaded System For Responding In Verse Authors: Devangi Sharma, Sophia Judicke, Glenda Tan, Conrad Schaumburg, Shinji Watanabe Published: 2026-09-20 Categories: cs.CL, cs.AI, cs.SD arXiv: https://arxiv.org/abs/2609.23951v1 Expressive speech synthesis has advanced through prosody modeling, yet generating structured poetic speech, such as haiku, remains challenging. Prior work on prosody transfer improves expressiveness, and fine-tuned poetry TTS (text-to-speech) systems capture verse intonation. However, these models do not model haiku's 5-7-5 syllable structure or line-ending pauses. We present a cascaded system, HaikuS2S, combining ASR (automatic speech recognition), LLM (large language model)-generated haiku, and TTS fine-tuning on both prose and custom haiku datasets. Our evaluation focuses on emotion similarity, speech quality, and prosody alignment. In our experiments, we see that our prosody and tonal alignment improve significantly with our fine-tuned systems, particularly the one trained on both general poetry and haiku. We also see that we maintain similar emotion similarity scores across all systems. --- ### ORION-CMR: On-scanner Reporting with Integrated Foundation Model for End-to-End Cardiac MRI Analysis and Interpretation Authors: Omer Burak Demirel, Kelly K. Horst, Alessio Perazzolo, Elisa Bruno, Kenan Kaya, Rongzhen Ouyang, Enas Ahmed, Jouke Smink, Spencer L. Waddle, Zainudeen Kallumpurath, Tzu Cheng Chao, Dinghui Wang et al. Published: 2026-09-20 Categories: eess.IV, cs.CV, cs.LG, physics.med-ph arXiv: https://arxiv.org/abs/2609.23950v1 Cardiovascular magnetic resonance (CMR) provides comprehensive cardiac assessment but remains underutilized because of the complexity of acquisition, post-processing, and interpretation. Existing artificial intelligence (AI) methods address isolated tasks, limiting clinical integration. We present ORION-CMR (On-scanner Reporting with Integrated fOunda-tioN Model), the first clinically evaluated scanner-native end-to-end CMR foundation model. Pretrained on 12,896,733 CMR images from 9,258 studies, ORION-CMR performs sequence classification, ventricular function assessment, late gadolinium enhancement (LGE) detection, binary and multiclass disease classification, and local large language model-based report generation in approximately 90 seconds. The framework. was evaluated on public benchmarks and clinically validated in a multi-vendor cohort of 68 subjects with normal examinations, congenital heart disease, dilated cardiomyopathy, and myocardial infarction. ORION-CMR outperformed supervised baselines and the previously published CMR foundation model (CMR-FM), achieving state-of-the-art performance for LGE classification and scar segmentation. Clinical evaluation achieved an AUC of 0.96 for normal-versus abnormal classification and 0.88 for multiclass disease classification, while generated reports demonstrated 81.4% agreement with expert interpretation. These results demonstrate the feasibility of real-time scanner-native AI-assisted CMR analysis and automated report generation. --- ### Echo State Network (ESN) for Signal Recovery in RF-Impaired IBFD MIMO Systems Authors: Conrad Prisby, Siyao Li, Chengtao Xu, Thomas Yang Published: 2026-09-20 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.23945v1 In-band full-duplex (IBFD) multiple-input multiple-output (MIMO) systems enable simultaneous transmission and reception on the same frequency band, improving spectral efficiency for next-generation wireless networks. However, IBFD-MIMO systems are susceptible to self-interference (SI), which may overpower signals of interest (SOI). In this scenario, blind source separation (BSS) algorithms can be adopted to remove SI and perform joint sensing and communication (JSAC), but BSS algorithms mostly assume an idealized linear and quasi-stationary signal model, which does not hold under realistic radio frequency (RF) impairments, such as I/Q imbalance, carrier frequency offset (CFO), phase noise, and power amplifier nonlinearity. This paper proposes a two-stage echo state network (ESN)-based scheme that is superior to BSS under these realistic conditions. A frozen ESN is trained offline to characterize the static SI path, while an adaptive ESN, updated online via recursive least squares, tracks the time-varying SOI path using sparse pilot symbols. We evaluate the proposed scheme's SOI recovery performance and acquisition speed with different block sizes, comparing it against other recurrent neural networks (RNN), such as long short-term memory (LSTM) and gated recurrent unit (GRU). Simulation results show that the proposed approach outperforms BSS, LSTM, and GRU in both efficiency and SOI recovery, demonstrating the viability of ESNs for real-time, nonlinear self-interference cancellation in realistic IBFD MIMO systems. --- ### XYEval: Agents say yes to bad advice Authors: Zhengxuan Wu, Yuxuan Li, Oyvind Tafjord, Been Kim Published: 2026-09-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.23939v1 Effective communication between users and AI agents is essential for human-AI collaboration. The XY problem is a well-known communication pitfall where a person asks about their attempted solution rather than their actual problem. We extend prior sycophancy evaluation to the XY problem in agentic settings, evaluating whether agents can resist plausible but misleading suggestions from users and communicate their reasoning. We introduce XYEval, a meta-evaluation framework that can transform an existing benchmark into an XY problem evaluation. We evaluate five models across six diverse benchmark suites. Agents suffer large XY drops under XY mutation across benchmarks, with relative drops reaching up to 46.7%. With $τ^2$-bench, we further show that agent performance drops more when encountering a pedantic user who requires detailed explanations before approving a better solution. Our findings suggest that current agents lack the ability to effectively reason and communicate when facing misleading suggestions. A simple system instruction baseline that encourages awareness of XY problems only offers partial mitigation. Extensive trace analyses provide behavioral insights into how and why these XY drops occur across execution trajectories. Our results show that mitigating the XY problem remains challenging, requiring agents to both recognize user misdirection and clearly communicate the underlying problem. --- ### Sparse Regression Distilled from a Single Robust Fit Authors: Wooyoung Shin, Seunghwan Park Published: 2026-09-20 Categories: stat.ML, cs.LG, stat.ME arXiv: https://arxiv.org/abs/2609.23937v1 Robust linear fits can resist response contamination yet remain too dense or unstable for useful global explanations. We propose penalized distillation, which fits a smoothly clipped absolute deviation (SCAD) estimator to a robust initial estimator's empirical fitted surface along a safeguarded coordinate-descent path and evaluates candidate states separately for fidelity, parsimony, perturbation stability, and held-out prediction. The new results attach to the states the algorithm actually computes. Conditional on a fixed uncontaminated design, deterministic bounds transfer response-replacement boundedness from the initial fit to every retained path state. Turning to fixed dimension, we characterize the oracle-support branch by its empirical-Gram projection and influence function, give conditions for covariance-weighted least-squares approximation equivalence, and establish a path-conditional generalized information criterion. By contrast, at large dimension-to-sample ratios the full-coordinate robust fit collapses without warning, and screening restores the construction. Under a sure-screening framework, the robustness bound and the support and selection guarantees transfer to the screened fit. Simulations separate robustness transfer from support recovery, efficiency, and computation across the dimension-to-sample ratio, with p up to 240, and the signal density, which isolates what the sparse stage adds once the screen over-selects. In a duplicate-grouped superconductivity study, the distilled estimator remains predictively stable under prespecified training-response shifts but retains 66.8--68.8 of 81 slopes. Stronger sparsification reduces the model to 12.6--14.0 slopes only at visible fidelity and prediction cost. Distillation therefore preserves predictive stability on these data without substantiating a compact coordinate-level explanation. --- ### Measuring the Assistant's Harmlessness Preferences on the User Turn Authors: Jord Nguyen Published: 2026-09-20 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.23935v1 Post-training turns a general next-token predictor into a chat model with a persistent assistant persona. If that persona is a character the model plays only on its own turns, its preferences should govern what the assistant says, not what the model predicts other speakers will say. We test this boundary and find that it does not hold: a safety-relevant preference of the assistant---for harmless over harmful tasks---shapes the model's predictions even on the user's turn, where the assistant is not the one speaking. We find that this preference is small or near-zero in pretrained base models, that it emerges through post-training, replicated across open-weight model families, grows with scale, and can be moved by narrow finetuning that never touches user turns. We claim that this is evidence that post-training does not merely install a shallow assistant persona, but instead generalises beyond just the local assistant turn, into the model's representation of the user. --- ### The Neural Forcing for Three-Dimensional Incompressible Navier-Stokes finite time blowup Authors: Beibei Li Published: 2026-09-20 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.23934v1 We present a two-part neural framework for forced three-dimensional incompressible Navier--Stokes flow. Part~I develops the computational forcing system. A physics-informed neural model generates structured external-force trajectories, candidates are optimized through differentiable PDE rollouts or PPO-Clip, and selected forcings are frozen and checked by independent fixed-force replay. Part~II provides the mathematical certification layer. It separates neural candidate discovery from continuum analysis, derives integrated reciprocal-vorticity criteria that imply Riccati-type growth and finite-time loss of smooth continuation, develops a validated computational-to-continuum transfer strategy, and establishes a conditional positive-probability closure for a nondegenerate neural output law. The proof is complete at the continuum level. --- ### Density-Ratio Rescoring for Imbalanced Classification Using Raking Duals and Classifier Scores Authors: Dongha Kim, Seunghwan Park Published: 2026-09-20 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2609.23926v2 Density-Ratio Rescoring (DRR) augments a classifier trained at the original class prior with a survey-raking dual score. Raking reweights the majority sample to match minority feature moments within a tolerance. DRR marginally standardizes the dual and base scores and combines them with a fixed weight of one half, using the fitted dual directly for prediction without resampling or refitting the base classifier. Under exact population matching and a correctly specified log-linear tilt model, the dual equals the log density ratio up to an additive constant. A class-separation analysis characterizes the signal strength and correlation conditions under which fusion improves separation under common within-class covariance. On 24 tabular benchmarks, evaluated over 30 trials and five base learners, DRR at the D=128 random-feature setting improves average precision over the standardized base on every dataset, with a mean gain of 0.034. It exceeds the shared-dual raking-and-relabeling resampler on 22 of 24 datasets, with a mean gain of $0.092$, and on all eight one-versus-rest tasks of a shared gene-expression cohort. These results demonstrate the effectiveness of using raking duals as reusable scores for improving rare-class ranking while retaining classifiers trained at the original prior. --- ### MCPGen: Benchmarking LLMs on Executable MCPWorkflow Development Authors: Yingxuan Yang, Jiaqi Liu, Lirui Guan, Jiaye Gao, Weiwen Liu, Weinan Zhang, Ying Wen Published: 2026-09-20 Categories: cs.SE, cs.AI arXiv: https://arxiv.org/abs/2609.23925v1 We study whether LLMs can produce executable workflow artifacts that remain consistent across graph structure, tool implementation, schema bindings, and runtime wiring. In this setting, correctness depends on cross-layer consistency: a workflow may be structurally plausible, yet still fail because tool implementations, schema bindings, or runtime execution do not align. Existing benchmarks largely evaluate these capabilities in isolation or rely on trajectory-level proxies, leaving open whether generated workflow artifacts execute end-to-end. We introduce \textbf{MCPGen}, an executable benchmark for Model Context Protocol (MCP) workflow development. MCPGen contains 100 self-contained MCP projects across 16 application domains and evaluates three diagnostic tasks: workflow reconstruction, tool creation, and backward-compatible workflow extension. We evaluate 11 representative LLMs in a single-turn foundation-model setting, assessing generated artifacts through static analysis, unit and integration tests, and process-isolated end-to-end execution. Models reach 88.5\% on workflow reconstruction, but no model exceeds 57\% end-to-end execution success. Per-tool unit-test pass rates reach 63.8\%, while project-level integration success does not exceed 45\%, suggesting that integration remains a major bottleneck even when isolated tool tests pass. --- ### Matched-Input Estimates Differ in Sign Across Architectures: Auditing EEG Foundation Models on Motor Imagery Authors: Kevin Zhou, Sparsh Roy Published: 2026-09-20 Categories: cs.LG, q-bio.NC arXiv: https://arxiv.org/abs/2609.23924v1 Pretrained EEG foundation models are increasingly proposed as general-purpose encoders for brain-computer interfaces, yet recent benchmarks disagree about when their representations transfer to downstream tasks. We audit LaBraM and CBraMod on motor imagery under a validation-locked protocol in which preprocessing, architecture, optimization, freeze depth, checkpoint, temperature, and method selection are determined using training-session data only. On four-class BCI Competition IV-2a, every supervised comparator evaluated here outperforms every foundation-model configuration, including validation-selected fine-tuning. We then examine a key confound: foundation models and task-specific decoders are normally evaluated using different input pipelines. Retraining three supervised architectures on the broadband arrays consumed by the foundation models produces matched-input accuracy differences of opposite sign across architectures: broadband input improves ATCNet by 0.078 accuracy while reducing EEG Conformer accuracy by 0.088. None of the three individual matched-input terms is significant after multiple-comparison correction at n = 9, so we treat the sign variation descriptively rather than as a formal architecture-by-pipeline interaction. These observed sign differences suggest that a single comparator may not provide an architecture-invariant decomposition of a pretrained-versus-supervised performance gap. The four-class deficit also does not reproduce uniformly across motor-imagery datasets: on two-class BNCI2014-004 we cannot detect the same separation between fine-tuned CBraMod and the supervised comparators. Finally, validation-fitted temperature scaling returns foundation-model calibration error to the supervised range despite substantially lower four-class accuracy. ---