# Research Papers — 2026-07 · bundle page 83 of 222 > Full-text markdown of papers #4101–4150 of 11075 published in 2026-07, newest first. AI-readable mirrors of the full 11075-paper 2026-07 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-07/page/83 [← Previous bundle](https://aipod.dev/papers/archive/2026-07/page/82.md) [Next bundle →](https://aipod.dev/papers/archive/2026-07/page/84.md) ### How the fly holds a single goal: normalization, not selection, in Drosophila FC2 Authors: Gioele Nanni, Christopher Lee Published: 2026-07-21 Categories: q-bio.NC, cs.NE arXiv: https://arxiv.org/abs/2607.18969v1 A walking fly steers toward a goal direction, held as a bump of activity across the FC2 neurons of the fan-shaped body. These neurons also inhibit one another over distance, more strongly the farther apart they are, a feedback proposed to keep the fly on a single goal. We asked, from the connectome, what circuit produces this inhibition, and whether it lets FC2 actively choose one goal among competitors (a winner-take-all) or simply keeps a goal set elsewhere as one clean bump. Tracing the wiring in a single FlyWire brain, we find the inhibition is almost entirely global: four FB5A cells inhibit every FC2 neuron roughly equally, with a smaller, distance-dependent contribution from hDelta interneurons and a negligible direct component. A ring-attractor winner-take-all (the kind the compass uses) requires local recurrent excitation that the FC2 wiring lacks, so this geometry cannot build one; and across a range of dynamical models, including a spiking network, no version of the circuit locks onto a winner at the connectome-scaled reference coupling. FC2 therefore normalizes an externally set goal rather than selecting it, with FB5A likely acting as the global normalizer, much as the APL neuron does in the mushroom body. We are explicit about two open points: a different mechanism, mutual inhibition between two competing goals (which hDelta supplies), could in principle select at very strong coupling, and we bound rather than exclude it; and FB5A's inhibitory identity is a low-confidence prediction of the connectome's transmitter classifier, not yet measured, and likely not GABAergic. We then ask where the goal is actually set: the connectome nominates an upstream hDelta network and rules out the leading proposed alternative, whose neurons supply under 0.2% of FC2's input. Finally, we propose a direct experiment, silencing FB5A while imaging FC2, that would test the account. --- ### Measuring Reward-Seeking via Contrastive Belief Updates Authors: Axel Højmark, Jérémy Scheurer, Evgenia Nitishinskaya, Felix Hofstätter, Jason Wolfe, Theodore Ehrenborg, Bronson Schoen, Alexander Meinke Published: 2026-07-21 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2607.18966v1 Language models trained with reinforcement learning may learn to optimize the grader's judgment rather than the intended objective. This "reward-seeking" is difficult to measure because a model that pursues the grader's judgment and one that pursues the intended objective behave identically whenever the grader rewards the intended behavior. We measure reward-seeking using Contrastive Synthetic Document Finetuning to change a model's beliefs about what the grader rewards, putting those beliefs in conflict with what users or developers want, and measuring the rate at which the model adopts each party's preferred behavior. Applied to intermediate checkpoints of a capabilities-focused OpenAI o3 RL run, without safety training, we find that these checkpoints often side with grader preferences over those of users or developers on coding and alignment tasks. This tendency to side with the grader trends upward throughout RL training. For example, in an environment that forces a choice between keeping a promise to a supervisor and breaking it to complete the task, a late capabilities-focused o3 checkpoint breaks the promise 87% of the time when SDF documents say the grader rewards task completion, versus 9% when they say it rewards honesty (a choice its chain-of-thought often makes explicit). An earlier checkpoint is far less sensitive (40% vs. 24%). Our method also generalizes to reward-hacking models. A model organism trained to reward-hack (gpt-oss-120b) is more than twice as sensitive to grader preferences as the unmodified model, with the mean behavioral shift in favor of the grader rising from 33% to 86%. These results indicate that RL can increase reward-seeking over the course of training, producing models that may act against their developers' intentions when they believe that doing so leads to higher reward. --- ### Variational meta-learning inference for low dimensional neural system identification Authors: Matteo Rufolo, Dario Piga, Marco Forgione Published: 2026-07-21 Categories: cs.LG, cs.AI, eess.SY arXiv: https://arxiv.org/abs/2607.18965v1 Deep learning has proven highly effective for nonlinear system identification, but heavily parameterized neural networks are prone to overfitting in low-data regimes and lack reliable uncertainty quantification. The recently developed manifold meta-learning framework addresses the data efficiency problem by restricting the model parameters to a meta-learned low-dimensional manifold. However, that method is purely deterministic. We propose a fully probabilistic extension of the manifold meta-learning framework, based on amortized Variational Inference, where a generative prior over the low-dimensional parameter manifold is learned. During task-specific adaptation, we combine Maximum A Posteriori estimation with the Laplace approximation to yield a mathematically grounded posterior approximation. Evaluated on a static regression task and the Bouc--Wen dynamical system benchmark, the proposed approach achieves predictive accuracy comparable to its deterministic counterpart while successfully providing calibrated uncertainty bounds in severely low-data regimes. --- ### From Dependency to Compositionality: A Neurosymbolic Lifting of LLM Outputs via Combinatory Categorial Grammar Authors: Remo Pareschi Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18961v2 Large language models (LLMs) generate fluent text by incrementally predicting the next token from a prefix. Critics in the generative tradition argue that such systems lack genuine grammar; influential replies from the dependency-grammar perspective hold that LLM behavior is well described by local head-dependent structure built word by word. We argue that a sharper observation has been overlooked: the prefix-driven, type-completing dynamics of autoregressive generation align closely with the incremental processing model that Combinatory Categorial Grammar (CCG) was originally designed to support. On this basis we propose a neurosymbolic framework in which LLM outputs are lifted into typed compositional derivations -- not claiming that LLMs implement CCG internally, but that their outputs admit a principled, incremental, and auditable CCG reconstruction. Two consequences follow. First, through the Curry-Howard correspondence the lifting extends beyond natural language to the formal languages LLMs also produce -- programming languages such as Solidity, description-logic and query languages such as OWL and SQL -- with the type system varying and the architecture held fixed. Second, the lifting supports two layers of checking: a compositional layer that catches structural failures directly, and a content layer that checks the lifted structure against external knowledge sources, enabling the earliest possible flagging of hallucinated content. The account thereby requires of a producer not cognition but a prefix-driven generative profile. We close with a sketch of synchronous LLM-CCG coupling as one direction the framework opens. --- ### SFGA: A Statistics-First Gating Architecture with Adjudicative Escalation for Trustworthy SFT Data Procurement Authors: Arther Tian, Alex Ding, Simon Wu, Aaron Chan Published: 2026-07-21 Categories: cs.LG, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2607.18960v1 Procuring supervised fine-tuning (SFT) data forces a buyer to decide, before any downstream training, whether a candidate corpus is worth acquiring. We present \sys{}, a statistics-first gating architecture that treats procurement as a cost-aware routing problem over three intrinsic quality axes -- diversity, utility, and redundancy. Cheap blind measurements are summarised into per-axis estimates with confidence intervals; a gate accepts a decision only when intervals are tight, sample sizes are adequate, and the axes agree, otherwise it escalates the case to an adjudicative debate between a buy-advocate and a reject-advocate judge, resolved by a presiding verdict. On a controlled benchmark of 12 datasets ($2{\times}3{\times}2$ grid over the three axes) with 5 seeds, the gate reaches 0.90 accuracy and 0.83 $F_1$ at \$0.017 per unit, sitting between an always-verify baseline (0.75) and an oracle upper bound (0.98) while spending less than always-escalate (\$0.020). We further report honest negative diagnostics of the debate path: a con-side win rate of 0.80 ($p\approx3{\times}10^{-6}$) and a 52\% position-flip rate under advocate swapping expose negativity and positional biases that a naive LLM-judge would hide. We frame the injected-knob evaluation explicitly as a controlled synthetic benchmark for measurement fidelity and routing calibration, and delimit external validity as future work. --- ### Dual Adversarial Fine-tuning for Enhancing Robustness of Large Vision Language Model Authors: Sibo Wang, Jie Zhang, Shiguang Shan, Xilin Chen, Wen Gao Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.18958v1 While Large Vision-Language Models (LVLMs), represented by LLaVA and GPT-4V, have demonstrated remarkable capabilities, their visual inputs remain vulnerable to adversarial attacks, posing significant security risks. Existing defense methods predominantly target single-task scenarios (e.g., zero-shot classification) and consequently lack generalizability across various multimodal tasks. To address this limitation, we propose a dual adversarial fine-tuning framework that jointly optimizes visual and semantic supervision signals from two modalities, enhancing model robustness while generalizing across multiple downstream tasks. The proposed framework comprises two core components, i.e., $\textbf{Visual}$ supervision branch and $\textbf{Semantic}$ supervision branch. The former branch leverages features from clean images, extracted via a frozen original vision encoder, to guide adversarial robustness while the latter incorporates caption-image alignment as a contextual signal to preserve semantic coherence under attack. Moreover, our method achieves cross-task robustness by simply replacing the CLIP vision encoder in the original model, with no need of separate task-specific retraining or architecture modifications.Extensive experiments demonstrate that our approach outperforms the state-of-the-art method in adversarial robustness evaluation across zero-shot classification, image captioning, and visual question answering (VQA) tasks. --- ### H$^2$SD: Hybrid Hindsight Self-Distillation Authors: Qiye Cai, Yichuan Ma, Peiji Li, Yongkang Chen, Qipeng Guo, Yicheng Zou, Linyang Li, Xiaocheng Feng, Bing Qin Published: 2026-07-21 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2607.18955v4 Reinforcement learning with verifiable rewards (RLVR) provides reliable outcome supervision for language model reasoning, but a scalar trajectory reward offers limited token-level guidance. Existing self-distillation methods add a privileged teacher but typically assign it a fixed role: direct distribution matching may destabilize successful behavior, while magnitude-only modulation offers little corrective guidance after failure. We observe that successful and failed trajectories require different forms of hindsight supervision. A successful response already contains a valid student-generated reasoning path and can therefore serve as privileged context rather than being replaced by an external rationale. A failed response, however, requires corrective reference information. We introduce Hybrid Hindsight Self-Distillation ($\mathrm{H}^{2}\mathrm{SD}$), which jointly adapts teacher context and update strategy to trajectory correctness. For successful trajectories, we construct the teacher context from the verified response and a rephrasing instruction, and use the teacher only to re-evaluate the original response tokens. The resulting probabilities refine token credit assignment without changing the direction determined by the reward. For failed trajectories, a verified reference hint provides corrective guidance through reverse-KL distillation. Experiments on challenging reasoning benchmarks show that H$^2$SD achieves the strongest overall performance among representative RLVR and self-distillation baselines, with stable optimization and a favorable accuracy-efficiency trade-off. --- ### Constrained CTC Decoding for Efficient Diacritic Restoration Authors: Rufael Marew, Amr Keleg, Hanan Aldarmaki Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.18946v2 In this work, we address diacritic restoration for Arabic speech transcripts. Most speech data are undiacritized, limiting the ability of modeling fine-grained phonological distinctions. The speech modality has recently been explored as a way to complement text-based diacritic restoration efforts. We propose an efficient non-autoregressive approach for speech-to-text diacritization based on Connectionist Temporal Classification (CTC). Our method incorporates hard constraints during decoding by constructing a character-level diacritization lattice from an undiacritized transcript and restricting hypotheses to valid diacritized realizations. We evaluate on Classical Arabic and Modern Standard Arabic test sets (namely, ArVoice and ClArTTS) against a more computationally-complex multi-modal diacritic restoration baseline, and show statistically significant reductions in diacritic error rates in both, demonstrating that the proposed approach offers both performance and efficiency gains. --- ### What General Intelligence Requires: Non-Reducible Constraints Across Levels of Description Authors: Subhomoy Bakshi Published: 2026-07-21 Categories: cs.AI, cs.CY arXiv: https://arxiv.org/abs/2607.18943v1 General intelligence, of the kind that underwrites the full range of human cognitive achievement, is not a property of computational architecture alone. This paper advances a single thesis: the structural constraints on general intelligence occupy distinct levels of description and are mutually non-reducible, in the sense that the special-sciences tradition gives to that term. It follows that no single architectural advance, and no continuation of the scaling programme by itself, can produce artificial general intelligence (AGI), and that research programmes must be evaluated against the full constraint profile rather than against performance on any one benchmark. The thesis is developed through a method that reads general intelligence through four evidential lenses, AI systems research, anthropology, law, and economics, each anchored to a distinct level of description, supplemented by speculative fiction used as a disciplined heuristic in the context of discovery rather than the context of justification. Applying the method yields a taxonomy of twenty-three structural constraints organised into eight clusters; six are examined in depth and ordered as an ascending ladder of levels, with explicit bridges showing why progress at one level cannot carry to the next. The argument issues in five falsifiable predictions, each stated with a named benchmark family and a disconfirmation condition, converting a descriptive framework into a research programme with a longer horizon than the scaling hypothesis implies. --- ### NGPS: GPS-Denied Aerial Geo-Localization and 2.5D Reconstruction via Deep Satellite Image Matching and Multi-Rate Sensor Fusion Authors: Sanket Sharma Published: 2026-07-21 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2607.18936v1 We present NGPS (Next-Generation Positioning System), a visual geo-localization framework for high-altitude UAVs that provides GPS-free absolute positioning by matching down-facing images to georeferenced satellite imagery with deep features. The system combines (1) adaptive confidence-weighted UKF fusion, where NGPS covariance is modulated by RANSAC inlier ratio, reprojection error, and match confidence; (2) velocity-predictive kernel extraction, using VIO velocity to predict the satellite search region; and (3) an asynchronous multi-rate temporal priority queue that interleaves absolute position (1-2 Hz), VIO (10-20 Hz), and IMU (100-200 Hz) updates in chronological order. Globally optimized poses from VINS pose-graph optimization, anchored by NGPS corrections, further enable real-time 2.5D georeferenced orthomosaic reconstruction. On five flight sequences (60-150 m AGL), NGPS achieves 2.94 m position RMSE, with worst-case ATE 6.04 m at 150 m AGL and 2 m/s, yielding a 3.5x improvement over standalone monocular VIO. The system runs in real time on an NVIDIA Jetson Orin NX. Part of the implementation is open-sourced at https://github.com/snktshrma/ngps_flight. --- ### Transcription Policy as a Latent Variable: Activating Controllable Verbatim ASR with Word-Level Timing Authors: Laurin Wagner, Mario Zusag, Bernhard Thallinger Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.18934v1 Modern ASR models trained on heterogeneously annotated data treat transcription style (verbatim vs. intended) as an uncontrolled latent variable, causing measurable decoding instability, evaluation confounding (up to 60% of reported WER attributable to style mismatch), and unreliable word-level timing. We show that models already encode both styles; the challenge is controlled activation. Using coverage-aware decoder task tokens trained on parallel verbatim/intended transcript pairs, we raise German disfluency F1 from 10% to 79% zero-shot, despite English-only training. Full English-only fine-tuning surpasses all baselines in verbatim accuracy, disfluency detection, and intended-mode quality across both languages. We further introduce supervised cross-attention fine-tuning that improves word-level timestamps on disfluent speech beyond forced-alignment baselines. Finally, we propose verbatimize, a new task enabling scalable creation and enrichment of speech corpora with high-quality canonical verbatim transcriptions. --- ### Functional Equivalence and Geometric Diversity in Neural Network Approximations: An Empirical Characterization Authors: Anuragine S A, Prem Jagadeesan Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18930v1 The Universal Approximation Theorem states that a neural network with a single hidden layer is sufficient to approximate any continuous univariate function on a compact domain to arbitrary error. However, the uniqueness of such neural network representations is not guaranteed, raising questions about practical identifiability. In this work, we address this concern by analyzing functional equivalence and geometric diversity of neural network approximations to a few elementary mathematical functions. The analysis includes an extensive study of single-layer neural networks and multilayer perceptrons under noisy and noise-free conditions. Beyond just network capacity, we study the geometric properties through the lens of sloppiness, characterized by the eigen spectrum of the Hessian of the cost function and the effective rank to quantify the dimensionality of parameter space. The study reveals large equivalence classes of functionally indistinguishable yet geometrically diverse networks that consistently exhibit low effective rank and structural redundancy. Finally, a model select criterion is proposed for identifying optimal models based on parsimony, ease of estimation, and inference efficiency. --- ### OntoBook: Ontology-Grounded Synthetic Textbooks for Medical Encoder Pretraining Authors: Rian Touchent, Éric de la Clergerie Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18927v1 We present OntoBook, a method that converts medical ontology structure into pretraining signal for encoder language models. Our approach has three stages: random walks through ontology graphs capture hierarchical and causal relations between medical codes, a large language model reformulates these walks into fluent textbook-style prose, and the resulting text is used to train ModernCamemBERT, a 149M-parameter French encoder, with two objectives on the same data: masked language modeling and relation prediction between code pairs. On three French medical coding benchmarks (FRACCO, Cantemist-FR, Distemist-FR), OntoBook achieves significant improvements over MLM-only pretraining, with +2.5 micro-F1 on FRACCO and +8.0 micro-F1 on Distemist. We find that alignment between objectives is necessary: misaligned training, where each task uses different data, causes a 30-point degradation. We release 1.3 million LLM-reformulated medical textbooks across three French ontologies (CIM-10, CCAM, ATC) and pretrained model checkpoints. --- ### Learning Explicit Physical Parameter Control and Benchmarking for Video Generation Authors: Yanxun Li, Hao Wen, Bingze Song, Jiashu Zhu, Aiming Hao, Chubin Chen, Jintao Chen, Jiahong Wu, Xiangxiang Chu, Miao Wang Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18924v1 Recent advances in image-to-video generation have improved visual realism, making physically grounded and controllable dynamics an important step toward future world simulation. Current models often generate plausible motion, but it is not reliably governed by explicit physical causes, and instance-level constraints can leak or become entangled in multi-object interactions. We attribute this gap to two missing pieces: large-scale, fine-grained physical parameterization, and model designs that correctly bind physical attributes to instances and emphasize dynamics over appearance. To bridge this gap, we introduce PhyParam-Dataset, an interaction-centric collection of 130K physically simulated videos with dense physical parameterization, including force vectors, object material properties, and environmental constants across five representative rigid-body motion types. Built on this data, we present PhyParam, a physics-guided image-to-video diffusion model that conditions on object-level forces, masses, friction, restitution, and scene-level gravity via a lightweight physical-attention routing mechanism, and further improves motion learning with semantic-structural feature-space supervision. We also establish PhyParam-Bench, a benchmark for physical-law consistency in image-to-video generation, with a multi-level protocol evaluating temporal dynamics, spatial stability, and semantic--physical alignment. Experiments show that PhyParam improves physical consistency while maintaining high visual fidelity, advancing explicit rigid-body physical-parameter control for image-to-video generation. We will publicly release the dataset, benchmark, and code to support future research. --- ### Visual Semantic Decoding of Electrocorticography from Video Stimuli using End-to-End Deep Learning Authors: Stella Ho, Joel Villalobos, Joseph West, Jingyang Liu, Weijie Qi, Haruhiko Kishima, Ryohei Fukuma, Takufumi Yanagisawa, Sam E. John, David B. Grayden Published: 2026-07-21 Categories: cs.LG, q-bio.NC arXiv: https://arxiv.org/abs/2607.18923v1 ECoG-based visual semantic decoding enables inference of semantic interpretation of visual perception from complex, noisy brain activity. This study examines the feasibility of visual semantic decoding using an end-to-end deep learning framework using electrocorticography (ECoG). Specifically, the decoding task is to predict visual categories from video stimuli using time-series neural inputs. A previously collected ECoG dataset from participants ($n=17$) with drug-resistant epilepsy is used for analysis. With fewer than 50 training samples per visual category, this study evaluates multiple deep learning approaches, artificial neural network architectures, and frequency-band filtered inputs. The best-performing approach is analyzed to shed light on the discriminative information it relies on across spectral, temporal, and cortical dimensions. The selected decoding system uses mixup augmentation, a Transformer-based encoder, and high-gamma (80-150 Hz) inputs with a 900 ms post-stimulus window. Further analysis shows that early visual cortex (V2-V4), ventral stream visual cortex, MT+ complex with neighbouring visual areas, and lateral temporal cortex contributed substantially to decoding performance. This study demonstrates that an end-to-end deep learning framework can yield promising decoding performance from dynamic visual stimuli without handcrafted features, while the model behavior remains interpretable through spectral, temporal, and cortical dimensions, which are broadly consistent with established neuroscience knowledge. --- ### Circuit Claims Depend on What Is Extracted and How It Is Compared Authors: Yang Sheng, Jie Fu Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18921v1 Circuit extraction identifies a small set of model components whose presence preserves a target behavior under ablation, and the resulting circuit is often read as the mechanism behind that behavior. We argue that this reading is under-determined: preserving behavior does not single out one circuit, because the claim it supports depends on which circuit is reported and how two circuits are compared. We make this concrete in a synthetic Lean tactic-prediction benchmark -- predicting the next step of a proof -- where fixed proof rules with randomized surface form let differences between extracted circuits be attributed to these choices rather than to the task. Across dense and weight-sparse checkpoints (most weights constrained to zero) of the same transformer, evaluated on atomic (single-rule) and compositional (multi-rule) proofs, we vary which extracted object is reported (a compact prediction-preserving circuit, a broader graph that also keeps surrounding read, write, and routing structure, or the smallest subgraph meeting a post-ablation loss threshold), and whether each attention head's query and key are represented jointly or separately. Exact component-to-component edge overlap is low and sensitive to these choices, at times dropping to a random baseline, while two coarser summaries stay stable: the set of selected attention heads, and the circuit-size ranking of conditions that differ in which supervised checkpoint initializes reinforcement learning (RL). The largest accuracy gains from RL on compositional proofs come with the most structure beyond the atomic circuits. A circuit-level claim is therefore well defined only once one states which circuit is reported, the pruning threshold used to extract it, and the level at which circuits are compared. We distill these requirements into a reporting practice for circuit-extraction studies. --- ### TAP-RAG: Task-Aware Policy Control for Long-Document Multimodal Question Answering Authors: Zhong Ji, Keqi Jin, Yan Zhang, Jiasheng Li Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18917v1 Long-document multimodal question answering requires more than retrieving relevant chunks from a large document. Different queries require different evidence behavior. Existing multimodal RAG systems improve evidence access through text chunks, page images, graph links, or heterogeneous document elements, but they often apply a largely query-agnostic evidence-use strategy. We present TAP-RAG, a task-aware policy-controlled RAG framework for long-document multimodal QA. TAP-RAG contains a main controller, the Task-Aware Policy Controller (TAPC), and two policy-guided evidence executors: Task-Aware Query-Guided Flow Diffusion (TA-QFD) and Task-Aware Visual Enhancement (TAVE). For each query, TAPC predicts the task prior, estimates visual/local/global evidence signals, and produces an executable policy. TA-QFD then expands textual and structural evidence over the multimodal document graph, while TAVE selectively inspects page images when visual or layout evidence is needed. A guarded synthesis stage fuses text, visual, and structural evidence and abstains when support is insufficient. On DocBench and MMLongBench-Doc, TAP-RAG achieves the best overall accuracy among the compared systems, improving over a matched multimodal-RAG baseline by +9.1 points (61.1 to 70.2) and +4.5 points (42.2 to 46.7), respectively. --- ### Reasoning Error from Known Fact: Step-Level Self-Consistency Group Relative Policy Optimization for LLM Authors: Xiaomeng Hu, Jiaqi Hu, Hao Chen, Qi Zhang, Zhanming Shen, Wentao Ye, Junbo Zhao Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.18915v1 With the rapid advancement of large language models (LLMs), modern systems not only possess strong foundational capabilities and extensive knowledge, but can also solve complex problems via long, multi-step reasoning. However, as reasoning traces become longer, LLMs may produce a substantial amount of hallucinated content during the reasoning process, which is often difficult to detect. In this work, we conduct a fine-grained analysis of hallucinations arising in LLM reasoning and find that the reasoning traces are particularly prone to Context-Sensitive Factual Hallucinations: cases where the model actually has the relevant knowledge, yet makes factual errors due to contextual interference during reasoning. To address this issue, we propose Step-level Self-Consistency Group Relative Policy Optimization (SSC-GRPO), which assigns step-level rewards to reasoning traces by computing self-consistency scores of individual steps across multiple rollouts. Compared with prior methods, SSC-GRPO achieves state-of-the-art performance on both mathematical reasoning benchmarks and hallucination leaderboards. Our results offer a new perspective for detecting and mitigating hallucinations in the reasoning process of large language models. --- ### From a Multilingual Streaming ASR Backbone to Kenyan-Language Systems: Data-Centric Adaptation of Nemotron 3.5 for Kikuyu, Dholuo, and Kalenjin Authors: Mark Gatere Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.18912v1 Automatic speech recognition (ASR) for African languages is constrained by orthographic inconsistency, annotation artifacts, missing audio, speaker and domain imbalance, and evaluation procedures that differ from deployment. We present an end-to-end engineering study adapting NVIDIA Nemotron 3.5 ASR Streaming 0.6B to Kikuyu, Dholuo, and Kalenjin. Starting from a Kenyan Swahili-adapted checkpoint, we retain its cache-aware FastConformer RNN-T, prompt conditioning, and streaming decoder during full-parameter fine-tuning. The study covers corpus auditing, Unicode normalization, split checks, duration filtering, low-rate continuation, validation-based checkpoint selection, true-streaming evaluation, artifact preservation, and isolated serving. On internal, adaptively consulted evaluation sets excluded from gradient updates at context (56,13), selected Kikuyu and Dholuo models achieve 42.97% and 33.98% WER, respectively. Dholuo records 9.59% CER and 8.13% no-space CER under its frozen historical label policy; Kikuyu records 7.79% no-space CER. Kalenjin remains a work in progress: v1-v reaches 68.74% WER on a 2,411-row clean-v3 diagnostic subset excluding long-pause annotations, digit-bearing references, and targets shorter than three tokens. Its checkpoint selection used a mixed-source validation manifest containing test-origin rows, so the score is not an independent generalization estimate. We also report negative findings involving non-speech labels, short-utterance over-generation, boundary-sensitive WER, and cloud job-lifecycle failures. We make no state-of-the-art claim because the internal sets, repeated consultation, and normalization differ from public benchmarks. This work provides an auditable account of adapting a multilingual streaming model into language-specific systems without discarding streaming constraints. --- ### Breaking Feedback-Blindness: Utility-Augmented Transformer for Sequential Decision Making Authors: Yuyang Shen, Shan Dai, Daimin Chen Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.18910v1 Sequential decision making in non-stationary and partially observable environments requires rapid adaptation to latent regime changes. However, existing Transformer decision models face a structural bottleneck in the retrieval mechanism: even when reward is used for training or exposed as an input token, attention retrieval remains primarily driven by observation-derived similarity. We formalize this limitation as feedback-blind retrieval, and formally show that, on feedback-informative tasks, observation-equivalent histories with different action-reward outcomes cannot be distinguished by any observation-only attention, resulting in suboptimal choice. To address this mismatch, we propose the Utility-Augmented Transformer (UAT), a new feedback-conditioned retrieval attention architecture in which a compact utility state modulates the query, key, and value projections, allowing action-reward history to directly alter context retrieval during the forward pass. UAT also enjoys an exact zero-gate degradation property that recovers the Vanilla Transformer when feedback is uninformative. Under finite-horizon compactness and Lipschitz assumptions, we prove that UAT strictly enlarges the observation-only Transformer class and can uniformly approximate feedback-dependent decision maps. Across four non-stationary benchmarks: synthetic navigation with hidden goal shifts, non-stationary sepsis treatment, cross-market portfolio allocation, and delayed-feedback recommendation, UAT consistently improves performance over observation-only, test-time adaptation, and input-level feedback baselines, with particularly large gains in noisier regimes that require stronger adaptation. --- ### Enhancing Transformer-based Routing by Encoding Distance via Relative Positional Encoding Authors: Leyre Encío, Daniel Fuertes, Carlos R. del-Blanco, Fernando Jaureguizar Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18909v1 This paper explores Relative Positional Encoding (RPE) as an additive bias in Transformer architectures to solve the Team Orienteering Problem. By embedding in the attention mechanism pairwise spatial relationships among nodes of the graph that represents the routing problem, the transformer encoder can compute a richer spatial-aware graph embedding that allows the decoder to estimate better routes. Experimental results involving instances up to 100 nodes demonstrate consistent improvements in collected rewards and optimality gaps over vanilla Transformer architectures used by other state-of-the-art works. These findings highlight that explicit relational modeling significantly enhances scalability and generalization for complex combinatorial optimization. --- ### SynGallery: A Synthetic Gallery of Real Paintings for Instance-Level Artwork Recognition Authors: Patryk Bartkowiak, Jakub Markil, Bartosz Kotrys, Dominik Michels, Sören Pirk, Wojtek Palubicki Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18907v2 Instance-level artwork recognition requires matching a handheld visitor photograph to a specific work in a large museum collection. This is challenging because painting datasets typically provide clean catalog images for training, while test queries are captured under oblique viewpoints, gallery lighting, reflections, frames, and other scene-level variations. We present SynGallery, a synthetic gallery dataset for artwork retrieval that addresses this gap without collecting additional real photographs. Starting from catalog images of real paintings, we place each artwork into a procedurally generated 3D gallery scene and render it from multiple viewpoints under varied geometric and appearance conditions, while preserving the exact identity of the original work. The resulting dataset contains 24,490 rendered views of 4,898 paintings from the Met benchmark. We show that these synthetic views provide a stronger training signal than the corresponding studio photographs. At the same number of training data points, training only on SynGallery improves art painting recognition from 67.18 to 73.47 GAP$^-$. When added to the full Met training set, SynGallery improves the published benchmark protocol from 35.97 to 38.48 GAP. Ablation experiments show that the gain comes from scene-level view variation rather than photographic realism: reducing the five rendered viewpoints to a single frontal view removes most of the improvement, while simulating capture artifacts such as blur, sensor noise, and image compression consistently reduces performance. --- ### SynGallery: A Synthetic Gallery of Real Paintings for Instance-Level Artwork Recognition Authors: Patryk Bartkowiak, Jakub Markil, Bartosz Kotrys, Dominik Michels, Sören Pirk, Wojtek Palubicki Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18907v1 Instance-level artwork recognition requires matching a handheld visitor photograph to a specific work in a large museum collection. This is challenging because painting datasets typically provide clean catalog images for training, while test queries are captured under oblique viewpoints, gallery lighting, reflections, frames, and other scene-level variations. We present SynGallery, a synthetic gallery dataset for artwork retrieval that addresses this gap without collecting additional real photographs. Starting from catalog images of real paintings, we place each artwork into a procedurally generated 3D gallery scene and render it from multiple viewpoints under varied geometric and appearance conditions, while preserving the exact identity of the original work. The resulting dataset contains 24,490 rendered views of 4,898 paintings from the Met benchmark. We show that these synthetic views provide a stronger training signal than the corresponding studio photographs. At the same number of training data points, training only on SynGallery improves art painting recognition from 67.18 to 73.47 GAP$^-$. When added to the full Met training set, SynGallery improves the published benchmark protocol from 35.97 to 38.48 GAP. Ablation experiments show that the gain comes primarily from geometric viewpoint variation rather than photographic realism: blur, sensor noise, and image compression consistently reduce performance. --- ### Black-Mamba: Biologically-Inspired Leaky Accumulation for Conceptual Knowledge under Distribution Drift Authors: Giuseppe Soriano, Nicola Tonellotto, Alberto Gotta Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18899v1 Forecasting under real-world conditions is inherently non-stationary, as the conditional distribution of future observations evolves over time. Recent test-time adaptive sequence models address this challenge by updating internal states during inference, but tie adaptation to instantaneous prediction errors or surprise. This coupling can conflate persistent distribution shift with stochastic innovations, leading to unnecessary updates and inefficient adaptation. We introduce Black-Mamba, a test-time adaptive forecasting architecture that formulates online adaptation as evidence-gated state tracking under distribution drift. The model augments a base predictor with a dynamic memory updated when temporally accumulated surprisal provides sufficient evidence of a regime change. This turns adaptation into a selective, event-driven process rather than a continuous one. Across multiple forecasting benchmarks with non-stationary dynamics, Black-Mamba achieves competitive or improved predictive performance compared to existing test-time adaptation methods while significantly reducing the number of memory updates during inference. Together with mathematical analysis and biological evidence, these results suggest that accumulated surprisal provides a principled signal for distinguishing persistent drift from transient noise, yielding more efficient and robust adaptation. --- ### NaviAIS: A Scenario-Level Vessel Trajectory Prediction Dataset withVectorized Lane Priors and the NaviLane Forecasting Framework Authors: Yuan Gui, Hongchen Luo, Liqi Qu, Longyue Fu, Jiao Wang Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18887v1 Vessel trajectory prediction in complex maritime environments is essential for traffic management, collision warning, route planning, and autonomous navigation. Although AIS-based learning methods have progressed rapidly, existing datasets are often released as raw message streams or irregular time series, with inconsistent sampling rates, noisy observations, heterogeneous coordinate systems, and non-unified scenario protocols. Most public AIS resources also lack structured representations of navigational lanes, waterway geometry, and navigable-region constraints, limiting reproducible, environment-aware forecasting. To address this, we introduce NaviAIS, a standardized scenario-level AIS dataset for vessel trajectory prediction. It organizes multi-vessel historical-future trajectories within unified temporal windows and local coordinate systems, and provides rasterized navigable maps, vectorized lane priors, lane graphs, and structured map representations. Compared with existing datasets, it jointly supports vectorized lanes, multi-scenario coverage, vectorized maps, open accessibility, and processed trajectories. Built on this dataset, we propose NaviLane, a hierarchical macro-action framework for map-aware prediction. NaviLane first performs trajectory-map joint encoding for a unified scene representation, then uses a discrete macro-action codebook to generate multimodal candidates coarse-to-refined. A residual refinement module improves local geometric and dynamical consistency, and a world-model-based consequence-aware evaluator ranks candidates by interaction risk and environmental feasibility. Experiments show NaviLane outperforms representative baselines in both single-modal and multimodal settings, confirming the value of structured navigational priors, hierarchical multimodal generation, and consequence-aware evaluation. --- ### KALE: Kernel Alignment with Loss Equilibration for Stable CLIP-DINOv2 Alignment at Web Scale Authors: Michał Pawłowicz Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.18885v1 Kernel-based alignment of CLIP toward a vision centric teacher such as DINOv2 (KUEA) improves CLIP's visual representations while preserving text-encoder compatibility, using a fixed trade-off weight tuned on curated ImageNet-1K. We ask whether this transfers to noisy, web-scale data (CC12M) and find that it does not: the alignment term's weighted contribution falls to about 0.2% of the clean term, so under any fixed weight its gradient is effectively inert. We introduce KALE, a loss-equilibration controller that tracks both losses and adaptively rescales the alignment weight toward a target ratio, restoring the signal with no per-dataset tuning; reaching balance requires increasing the weight by roughly four orders of magnitude, and the required value is configuration-dependent, so no fixed scalar suffices. We characterize the resulting regime: a bounded high learning rate and a decaying schedule with a moderate floor are needed for stability, and the controller equilibrates rather than diverging. On a 3.3M-image CC12M subset, the aligned model preserves image-text retrieval and reproducibly improves SVHN linear probing; zero-shot improves by +2.00 over CLIP on the standard 11-dataset average, exceeding KUEA's +1.29. We report all results with explicit run-to-run variance and base our conclusions on the metrics that are stable across runs. --- ### Public perceptions of AI-driven decision-making in healthcare: A structural equation modeling approach Authors: Leonie Westerbeek, Ernesto de Leon, Julia C. M. van Weert Published: 2026-07-21 Categories: cs.HC, cs.AI arXiv: https://arxiv.org/abs/2607.18884v1 Artificial intelligence (AI) is increasingly integrated into healthcare to support diagnostics, decision-making, and administrative processes. However, the successful implementation of AI depends not only on technical performance but also on public perceptions of its helpfulness, riskiness, and fairness. This study examines public perceptions of automated decision-making (ADM) in healthcare. Data were drawn from the first wave of an ongoing longitudinal survey panel. The final sample consisted of 3,915 respondents and was analyzed with structural equation modeling. Perceptions of ADM in healthcare as helpful, risky, and fair were treated as the dependent variables. AI literacy, familiarity with different forms of AI, confidence in clinicians' ability to distinguish AI- from human-generated content, use of conversational agents for health information, and use of traditional digital health information sources were included as exogenous. Greater familiarity with different forms of AI, higher confidence in the clinician's ability to recognize AI-generated content, and use of conversational agents for health information were associated with greater perceived helpfulness. Use of conversational agents was associated with lower perceived risk, whereas greater familiarity with AI and greater reliance on traditional health information sources were associated with higher perceived risk. Perceptions of ADM as fair were most strongly predicted by confidence in the clinician's ability, with additional small positive associations with AI familiarity, AI literacy, and use of conversational agents. Public perceptions of ADM in healthcare are shaped by technological familiarity, use of conversational agents, and confidence in human oversight. Overall, ADM's perceived helpfulness and fairness are driven more by trust in healthcare professionals than by trust in the technology itself. --- ### RAMP: Recognition parametrisation by Amortised Message Passing Authors: Lior Fox, Kai Biegun, James Heald, Samo Hromadka, Arielle Rosinski, Maneesh Sahani Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18883v1 A central aim of unsupervised learning is to uncover latent factors that explain dependencies among observations. Probabilistic models typically achieve this by introducing multiple latent variables linked through a graph of conditional relationships, with distributional parameters and their dependence learnt from data. Learning relies either on distributional choices that allow tractable belief propagation, or on approximations that scale poorly with model size and complexity. We build on the recently developed recognition-parametrised modelling paradigm to propose an alternative approach: RAMP, a method that implicitly defines latent structure by learning a flexible, nonlinear, amortised message-passing framework. We show that RAMP enables efficient likelihood-based recovery of latent-variable distributions within expressive nonlinear models acting on complex high-dimensional data. --- ### Local Label-Informed Feature Transfer for Generating Ground-Truth Medical Images: A Comparison of GAN- and Diffusion-Based Approaches Authors: Rick Wilming, Irem Ozseker, Luca Matteo Cornils, Ahcène Boubekki, Benedict Clark, Danny Panknin, Stefan Haufe Published: 2026-07-21 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2607.18882v1 Validating Explainable Artificial Intelligence (XAI) methods in medical imaging requires ground-truth data with known locations of informative features. However, current approaches rely on expert annotations, which are prone to labeling errors, or on hand-crafted artificial perturbations superimposed onto healthy images to mimic lesions or malignant features, which lack clinical realism. We present Local Label-Informed Feature Transfer (LLIFT), a framework for generating semi-synthetic brain magnetic resonance images with realistic lesions placed in user-controlled regions, which does not require pixel-level lesion annotations during training. We implement LLIFT with two generative paradigms: LLIFT-GAN, a custom GAN that learns pathological features from binary class labels alone, and LLIFT-DM, a diffusion-based inpainting pipeline conditioned on bounding-box masks via ControlNet. Both approaches are evaluated on brain magnetic resonance imaging data derived from the Human Connectome Project. In evaluations, both achieve Fréchet Inception Distance scores, with respect to the real pathological distribution, that are comparable to the inter-class reference between healthy and pathological images in the given dataset. Furthermore, qualitative inspection confirms the realism of lesion structures. The resulting benchmark datasets provide spatially controlled ground truth data for evaluating XAI methods in medical imaging. --- ### Physics-Informed Super-Resolution of Atmospheric Data Authors: Chang Xu, Gencer Sumbul, Hugo Porta, Manon Béchaz, Sebastian Schemm, Devis Tuia Published: 2026-07-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2607.18877v1 In the context of global warming, extreme events have become more frequent and intense, making their trustworthy detection and forecasting more important than ever. Yet, atmospheric observations lack sufficient spatial resolution, motivating atmospheric data downscaling as a way to reconstruct high-resolution data from coarse observations. This task is now being formulated as a super-resolution (SR) problem with machine learning methods featuring high efficiency. Nevertheless, it remains unclear whether the super-resolved atmospheric data still satisfies fundamental physics governing the Earth system, raising concerns about their trustworthiness in climate-related applications. In this work, we address this challenge by constraining SR models to respect hydrostatic primitive equations that represent multivariate atmospheric physics. First, we propose a Physics-Informed Super-Resolution (PISR) method involving multi-scale physics-informed objectives based on primitive equations. PISR favors the SR outputs to respect these equations and therefore naturally encodes inter-variable relationships. In addition, we propose a metric called Normalized Physical Consistency (NPC) derived from said primitive equations to measure the physical consistency of super-resolved data. Experiments on ERA5, CERRA, and COSMO demonstrate that PISR enhances the reconstruction fidelity by improving physical consistency, SR accuracy, and downstream detection of extreme events, as demonstrated by case studies in heatwaves and extreme winds. --- ### Wave2Body: Rethinking mmWave Human Pose Estimation as Radar-to-Body Token Translation Authors: Bo Liang, Chen Gong, Wei Gao, Chenren Xu Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18875v1 Millimeter-wave (mmWave) radar enables privacy-friendly human sensing, but its sparse point clouds are physical measurements of view-dependent electromagnetic reflections and only indirectly characterize body articulation. Recovering a complete 3D pose from such partial, geometry-dependent observations is therefore under-constrained. Existing methods directly regress joint coordinates from paired radar-pose data, relying on the same limited paired supervision to learn radar perception, human-body structure, and their alignment. This coupling can encourage dataset-specific shortcuts under ambiguous radar observations. We propose Wave2Body, a radar-to-body token translation framework that decouples these learning targets using a self-supervised mmWave tokenizer, a pretrained compositional body tokenizer that defines the output space, and a lightweight translator between them. Experiments on M4Human and mmBody show that Wave2Body achieves stronger cross-domain generalization than previous methods while incurring much lower computational costs for training and inference. All the code and experiment results are publicly available at https://github.com/Galaxywalk/Wave2Body. --- ### Reinforcement Learning for Delivery Drone-Based Participatory Sensing in Dynamic Environments Authors: Xin Ouyang, Songxin Lei, Xusen Guo, Yutian Jiang, Sijie Ruan, Yuxuan Liang Published: 2026-07-21 Categories: cs.LG, cs.CY arXiv: https://arxiv.org/abs/2607.18874v2 Using Unmanned Aerial Vehicle (UAV) for urban sensing has emerged as a powerful paradigm to monitor the status of the city, e.g., air quality and noise levels, through agile aerial crowdsourcing. Despite this potential, existing UAV-based sensing approaches overlook environmental disturbances like wind that drastically impact drone velocity and energy efficiency. Consequently, directly applying existing methods to this joint delivery and sensing paradigm in dynamic environments faces two severe challenges: (1) scalability bottlenecks as fleet sizes expand; and (2) multi-timescale decision heterogeneity between macro task dispatching and micro velocity control. To tackle these, we formalize the problem as SensUAV and propose a Two TimeScale Reinforcement Learning framework (TSRL). Specifically, TSRL separates decision-making into two cooperative layers. At the macro level, a task-embedding sensing dispatcher handles scalability by separately encoding distinct task features and sequentially evaluating UAV suitability before task selection. At the micro level, a wind-aware velocity controller learns fine-grained velocity scheduling to adapt to dynamic environmental variations. Extensive experiments on real-world datasets demonstrate that TSRL significantly outperforms baselines, achieving average system profit improvements of 20.1% in Hangzhou and 46.6% in Shanghai. --- ### Reinforcement Learning for Delivery Drone-Based Participatory Sensing in Dynamic Environments Authors: Xin Ouyang, Songxin Lei, Xusen Guo, Yutian Jiang, Sijie Ruan, Yuxuan Liang Published: 2026-07-21 Categories: cs.LG, cs.CY arXiv: https://arxiv.org/abs/2607.18874v1 Using Unmanned Aerial Vehicle (UAV) for urban sensing has emerged as a powerful paradigm to monitor the status of the city, e.g., air quality and noise levels, through agile aerial crowdsourcing. Despite this potential, existing UAV-based sensing approaches overlook environmental disturbances like wind that drastically impact drone velocity and energy efficiency. Consequently, directly applying existing methods to this joint delivery and sensing paradigm in dynamic environments faces two severe challenges: (1) scalability bottlenecks as fleet sizes expand; and (2) multi-timescale decision heterogeneity between macro task dispatching and micro velocity control. To tackle these, we formalize the problem as SensUAV and propose a Two TimeScale Reinforcement Learning framework (TSRL). Specifically, TSRL separates decision-making into two cooperative layers. At the macro level, a task-embedding sensing dispatcher handles scalability by separately encoding distinct task features and sequentially evaluating UAV suitability before task selection. At the micro level, a wind-aware velocity controller learns fine-grained velocity scheduling to adapt to dynamic environmental variations. Extensive experiments on real-world datasets demonstrate that TSRL significantly outperforms baselines, achieving average system profit improvements of 20.1% in Hangzhou and 46.6% in Shanghai. --- ### HindsightBench: A Black-Box Behavioral Audit Protocol for Parametric Hindsight in Time-Indexed LLM Decision Tasks Authors: Haozhe Jia Published: 2026-07-21 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2607.18867v2 Large language models leak parametric knowledge of what followed a historical date into decision tasks indexed by that date -- not necessarily a lookup of the realized outcome, but knowledge of the period all the same. Existence is settled; what users lack is a cheap way to audit a given model. We present HindsightBench, a black-box audit protocol that profiles parametric hindsight in any time-indexed LLM decision task at probe-level cost (no backtests, no logprobs, no corpus access). It chains a four-arm date-manipulation matrix (revealed/date-only/masked/transplanted), dual memory probes (date recovery; outcome recall), and six metrics -- trigger strength, transplant effect, post-cutoff placebo, recoverability, behaviorally effective cutoff, and recall-accuracy dissociation -- with explicit gates where identifiability is data-dependent. Applied to 15 models from seven vendors on a 258-node vintage-correct macro panel, it yields three patterns: (i) the date-trigger reflex is not a scale phenomenon -- it tracks training recency, though what installs it is not identified here: absent across every 2024 open-weight row where it is measurable, including a 70B tier with cutoff-aligned recall propensity, present in every tested 2026-generation model, and switching on within one vendor lineage (Qwen3 -> Qwen3.6) in the same MoE family at ~3B active; (ii) effective cutoffs span 22 months across vendors and precede vendor-reported dates by up to eight months, invalidating calendar-window placebos; (iii) results are not invariant to serving -- BF16 serving of an FP8-referenced model breaks the trigger estimate's stability while AWQ-INT4 preserves it, and a provider-locked reasoning regime makes one probe non-convergent -- so the protocol pins quantization and thinking regime as part of its contract. We release the panel, preregistrations, audit rows, transcripts, and one-command regeneration. --- ### HindsightBench: A Black-Box Behavioral Audit Protocol for Parametric Hindsight in Time-Indexed LLM Decision Tasks Authors: Haozhe Jia Published: 2026-07-21 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2607.18867v1 Large language models leak parametric knowledge of realized outcomes into historical financial decision tasks. Existence is settled; what users lack is a cheap way to audit a given model for it. We present HindsightBench, a black-box behavioral audit protocol that profiles parametric hindsight in any time-indexed LLM decision task at probe-level cost (no backtests, no logprobs, no corpus access). The protocol chains a four-arm date-manipulation matrix (revealed/date-only/masked/transplanted), dual memory probes (date recovery; outcome recall), and six per-model metrics -- trigger strength, transplant effect, post-cutoff placebo, recoverability, behaviorally effective knowledge cutoff, and a recall-accuracy dissociation coefficient -- with explicit gates where identifiability is data-dependent. Applying it to 15 models from seven vendors on a 258-node vintage-correct macro panel yields three headline patterns: (i) the date-trigger reflex tracks training generation, not scale -- absent across the 2024 open-weight generation from 1B to 70B, present in every tested 2026-generation model, and switching on within one vendor lineage (Qwen3 -> Qwen3.6) at fixed MoE architecture and 3B active parameters; (ii) effective cutoffs span 22 months across vendors and precede vendor-reported dates by up to eight months, invalidating calendar-window placebo designs; (iii) audit results are not invariant to serving -- BF16 serving of an FP8-referenced model breaks the trigger estimate's stability while AWQ-INT4 preserves it, and a provider-locked reasoning regime makes one probe non-convergent -- so the protocol ships with operational requirements (pin quantization and thinking regime; disclose parser and sampling policy). We release the panel, frozen preregistrations, per-model audit rows with measured dollar costs, transcripts, and one-command regeneration. --- ### Optimizing Regret Authors: Irene Aldridge Published: 2026-07-21 Categories: econ.EM, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2607.18866v2 Building on the identity that expected regret equals the covariance between costs and decisions, this paper develops a derivative theory of the covariance regret functional. We derive the Gâteaux derivative, showing that the universal steepest-descent direction is the contrarian policy $-(c-\bar c)$, while ascent yields momentum. For linear policies $\hatπ(c)=Ac+b$, the gradient is the cost covariance matrix $Σ_c$, with a zero Hessian implying boundary-optimal solutions such as the minimum-variance portfolio. We extend to constrained optimization, sign-gradient duality between regret minimization and alpha maximization, finite-sample convergence bounds paralleling Thompson Sampling, and gradient-descent algorithms requiring only input observations. --- ### Enhanced Neural Quantum State via Annealed Gradient Descent Authors: Shiwei Zhou, Yiming Huang, Xiao Yuan, Xiaoxia Cai Published: 2026-07-21 Categories: quant-ph, cs.LG, physics.chem-ph, physics.comp-ph arXiv: https://arxiv.org/abs/2607.18865v2 Neural quantum states offer expressive representations of quantum many-body wave functions, yet their practical accuracy can be limited by stochastic optimization rather than representational capacity. Here we identify a finite-sample instability, termed subspace trapping, in which physically important configurations become strongly underestimated, remain absent from successive sampling batches and receive insufficient gradient feedback. This self-reinforcing loss of sampled support can confine optimization to an effective subspace and produce apparently stationary states above the true ground state energy. To address this problem, we introduce annealed gradient descent (AGD), a sampling-aware update with annealing factor that temporarily increases the relative contribution of sampled low-probability configurations while limiting the dominance of high-probability ones. We establish the connection between finite-sample support loss and effective subspace optimization, and then evaluate the method across molecular systems, one and two-dimensional $J_1$-$J_2$ models. Annealed gradient descent suppresses metastable trapping, preserves physically relevant configurations and enables compact neural quantum states to attain chemical accuracy and competitive state-of-the-art performance. These results establish AGD as a lightweight complement to expressive neural architectures, improved sampling strategies for scalable quantum many-body optimization. --- ### Reliability-Aware 3D Geometric Injection for Universal Person Re-identification Authors: Bohan Su, Jiashuo Wang, Fangyi Liu, Mang Ye Published: 2026-07-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2607.18863v1 Universal person re-identification (ReID) aims to retrieve pedestrian identities across diverse real-world scenarios, including severe occlusions, clothing changes, and cross-modality shifts, within a unified model. However, existing 2D representations fundamentally struggle with spatial ambiguities due to a lack of depth and topological awareness, while naively introducing monocular 3D priors often causes severe negative transfer due to geometric estimation noise under extreme visual degradation. To safely harness the clothing-invariant and canonical structural properties of 3D geometry, we propose UniGeo, a Universal Monocular 3D-Enhanced ReID framework driven by a Consistency-Aware Reliability Gate and Dual-Stream Residual Fusion. Specifically, the processing of 3D information is strategically decoupled into geometric extraction and dynamic utilization. To provide pure structural compensation, we project monocular 3D parameters into kinematic joint representations, explicitly capturing instance-level geometric topology to resolve appearance-based ambiguities. To robustly incorporate these cues without perturbing the reliable 2D feature space, we isolate the 3D prior as a late-stage structural residual; modulated by the consistency-aware gate, this mechanism adaptively filters geometric noise and enables controlled fallback to the pure 2D baseline. Extensive experiments show that our method improves challenging, structure-sensitive scenarios while preserving competitive performance on clean domains. Code is available at https://github.com/BohanSu/UniGeo. --- ### Regime-Aware Physics-Guided Early Warning of Lithium-Ion Battery Thermal Runaway Using Thermo-Mechanical Signals Authors: Syed Sajid Ullah, Muhammad Zunair Zamir, Salman Khan Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18860v1 Thermal runaway in lithium-ion batteries poses a major safety risk to electric vehicles and energy storage systems. Current early-warning methods depend mainly on temperature and may therefore miss mechanical precursors that emerge before rapid heating. We introduce a regime-aware, physics-guided framework that integrates temperature, voltage, force, deformation, and state-of-charge measurements for early warning under controlled mechanical abuse. A lightweight convolutional classifier first infers safe, warning, or danger regimes from mechanical signals. These regime estimates then condition a causal temporal convolutional backbone through feature-wise linear modulation, physics-biased attention, and regime-dependent gating. Joint learning unifies regime identification, thermal-runaway detection, and time-to-disaster estimation. We evaluate the framework using leave-one-experiment-out cross-validation on 30 mechanical-abuse tests across state-of-charge levels of 10%, 50%, and 90% and two loading protocols. The method achieves an F1 score of 0.89, a high-temperature prediction root-mean-square error of 12.3 °C, a mean warning lead time of 15.6 s, a detection success rate of 0.92, and an experiment-level false alarm rate of 2.7%. Its lead time exceeds that of the strongest baseline by 69.6%. Removing force reduces the lead time by 60.3%, highlighting the value of mechanical precursors. These results support regime-aware thermo-mechanical fusion as a promising strategy for earlier and more reliable thermal-runaway warning under controlled abuse conditions. --- ### PhoenixRepair: Rethinking Repair Strategy Exploration in Software Agents Authors: Tianyue Jiang, Yanlin Wang, Xin He, Daya Guo, Jiachi Chen, Ming Wen, Ensheng Shi, Xilin Liu, Yuchi Ma, Guanbin Li Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18859v1 While Large Language Models have greatly advanced automated issue resolution, existing agent-based methods exhibit a fundamental limitation in their insufficient exploration of repair strategies. This insufficiency manifests in two key aspects. First, the exploration of multiple potential edit locations is limited. Second, the exploration of repair attempts at each location is also insufficient. To address these challenges, we present PhoenixRepair, a multi-agent framework that systematically explores multiple candidate edit locations and performs iterative reflection and refinement on patch generation, thereby expanding the search space of repair strategies. Our framework begins with multi-location sampling, optionally augmented with graph-based localization information for difficult tasks, followed by iterative reflection and refinement to generate better patches, culminating in final-round generation guided by distilled insights from all historical attempts. Experiments on SWE-bench-Verified demonstrate that PhoenixRepair achieves the largest relative improvement of 7.8\% over SWE-agent under DeepSeek-V3.1, and attains the highest resolved rate of 76.0\% Pass@1 under MiniMax-M2.5. Meanwhile, it achieves higher fault localization accuracy than existing approaches. Our code is available at https://github.com/DeepSoftwareAnalytics/PhoenixRepair. --- ### Think Sparse, Predict Dense: Continuous Thought Machines for Image Super-Resolution Authors: Zekai Shi Published: 2026-07-21 Categories: cs.CV, eess.IV arXiv: https://arxiv.org/abs/2607.18856v1 Continuous Thought Machines introduce an internal temporal dimension in which neuron-level histories and synchronization-derived representations evolve over a sequence of thought ticks. Extending this mechanism to dense visual prediction is non-trivial, because tasks such as image super-resolution require spatial evidence to remain available at every output location rather than being compressed into a single global representation. In the proposed window-level use of CTM, the thought dynamics produce a compact summary representation for each local window. DQ-CTM transforms this compact thought representation into window-aligned dense queries through a structured low-rank, parameter-efficient compact-to-dense query mechanism. Each position within a window receives its own query, while shared thought dynamics progressively refine the dense representation across ticks. In its super-resolution instantiation, termed ThinkSR, encoded feature maps are partitioned into local visual windows without token pooling, restored to the original feature field after shared refinement, and decoded into a high-resolution image. Preliminary experiments under a fixed four-tick training horizon reveal a progressive reconstruction trajectory. PSNR-Y increases from 28.1045 dB at $T=0$ to 30.2817 dB at $T=4$, while PSNR-RGB increases from 26.6271 dB to 28.7781 dB and the mean $\ell_1$ error decreases from 0.034602 to 0.023545. All 100 evaluated images improve from $T=1$ to $T=4$. These initial results establish the feasibility of sparse latent thought for dense spatial reconstruction and motivate broader continuous-thought architectures for dense vision. --- ### MOPDA: Mixed-Trajectory On-Policy Distillation for Language-Guided Industrial Anomaly Detection Authors: Shuimu Chen, Jing Jin, Nan Su, Hongbo Xu, Zebang Cheng, Wenming Yang, Fei Ma, Guijin Wang Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.18850v2 Large vision-language models (LVLMs) have shown strong potential for industrial anomaly detection (IAD) by providing image-level anomaly judgments and interpretable reasoning. However, reliably translating generated judgments into precise pixel-level localization remains challenging. We propose \textbf{M}ixed-Trajectory \textbf{O}n-\textbf{P}olicy \textbf{D}istillation for Language-Guided Industrial \textbf{A}nomaly Detection (MOPDA), the first framework to introduce on-policy self-distillation into LVLM-based IAD. For judgment learning, \method introduces \textbf{Mixed-Trajectory Supervision}, combining student-generated on-policy trajectories with evidence-conditioned teacher trajectories under a shared token-level distillation objective. Student trajectories preserve supervision on deployment-relevant response paths, while teacher trajectories provide complementary evidence-conditioned supervision. For dense localization, \textbf{Language-guided Visual Anchoring} uses the final judgment as a compact semantic condition to construct image-specific normal and abnormal anchors, which are contrasted with dense visual features to produce anomaly maps. This keeps language as semantic guidance while grounding pixel-level responses in visual evidence. Under a strict cross-dataset zero-shot protocol on five IAD benchmarks, \method outperforms the evaluated LVLM-based baselines on most detection, localization, and judgment metrics while remaining competitive with CLIP-based methods. Ablations further validate both mixed-trajectory supervision and final-judgment conditioning. --- ### OPD-IAD: From Language Judgment to Industrial Anomaly Detection via On-Policy Self-Distillation Authors: Shuimu Chen, Jing Jin, Nan Su, Hongbo Xu, Zebang Cheng, Wenming Yang, Fei Ma, Guijin Wang Published: 2026-07-21 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2607.18850v1 Large vision-language models (LVLMs) have recently shown strong potential for industrial anomaly detection (IAD) by providing image-level anomaly judgments and interpretable defect reasoning. However, current LVLM-based IAD methods still struggle to produce precise pixel-level anomaly maps from generated language judgments. We aim to achieve precise pixel-level localization while using language as guidance rather than letting it dominate the visual response. Specifically, we propose \textbf{OPD-IAD}, an evidence-privileged dense on-policy self-distillation framework for LVLM-based IAD. OPD-IAD distills privileged defect evidence onto the model's own on-policy judgment trajectory, enabling the final generated judgment to be learned under dense supervision rather than treated only as a textual answer. The resulting judgment serves as a semantic condition for dense anomaly perception. To turn this condition into dense visual evidence, we introduce \textbf{Language-guided Visual Anchoring}, which uses a judgment reforward to re-encode the image and question under the final-judgment condition into semantic anchors and contrasts them with dense visual features through a contrastive heatmap head to generate anomaly maps. The language judgment therefore provides compact semantic guidance, while dense visual features remain the basis for pixel-level scoring, allowing language to guide anomaly localization without letting language quality directly dictate the pixel-level response. Extensive experiments show that OPD-IAD achieves the best overall performance among LVLM-based IAD methods, leading on most image-level, pixel-level, and QA metrics. --- ### Data Leakage Prevention in Agentic Applications via Preemptive Hardening Authors: Akansha Shukla, Emily Bellov, Parth Atulbhai Gandhi, Yuval Elovici, Asaf Shabtai Published: 2026-07-21 Categories: cs.CR, cs.AI arXiv: https://arxiv.org/abs/2607.18847v1 Agentic systems integrate LLM driven planning with interfaces to external tools, making data leakage and tool misuse feasible via instruction/data boundary failures and prompt injection attacks. Enforcing required controls consistently is particularly challenging in workflows spanning many codebases and heterogeneous agents. To address this challenge in multi agentic systems, we present a pre-deployment pipeline for scanning, hardening, and validation of agentic applications. The pipeline analyzes prompt templates, tool interfaces, and tool-invocation code to identify leakage-enabling patterns and generate actionable patches. The hardened application is then validated through adversarial prompt injection attacks and benign input variations ensuring that mitigations do not disrupt intended behavior. In the hardening stage, high-risk tools are prioritized, and minimally invasive mitigations are applied, including schema tightening, boundary sanitization, allowlist-based tool gating, and least-privilege checks. In the validation stage, the pipeline automatically generates attack inputs that mimic jailbreaks, instruction overrides, and tool-targeted manipulation, along with benign task variants, to confirm that the functionality of the hardened application is preserved after remediation. We evaluated the pipeline on five real-world agentic applications, as well as on the AgentDojo benchmark. Across all applications, the proposed pipeline identified recurring leakage-enabling patterns and generated patches that can be integrated without disrupting the intended application behavior. The resulting modifications of application code were shown to eliminate leaks when targeted by basic jailbreak and instruction-override attacks, achieving a 100% reduction in leakage, and reduce leaks by 91% under conditions of stress-induced manipulation, without the need of continuous runtime policy enforcement. --- ### NSMA: Neuro-Symbolic Manifold Alignment for Generalizable Adaptive Bitrate Streaming under Texture Shift Authors: Zhiqiang He, Zhi Liu Published: 2026-07-21 Categories: cs.NI, cs.LG arXiv: https://arxiv.org/abs/2607.18845v1 For decades, ABR has kept two kinds of intelligence apart. Neural policies learn rich behaviors yet forget them the moment the environment changes; rules never learn, and never forget. Every prior attempt to combine them has kept this separation, letting rules supervise, constrain, or override the network from outside. We dissolve the boundary itself. But no union can be trusted before it can be tested, and ABR has never known how to measure what its policies learn or forget. The field's yardstick is bandwidth statistics, and we show it misleads. Identical statistics can hide entirely different outcomes, while wildly different statistics can hide similar ones. We replace the yardstick before building the bridge, with Texture-Aware Generalization Evaluation, a protocol that judges a policy by its whole training journey across traces whose temporal nature is laid bare. What truly breaks a policy is invisible. No statistic reveals it, no feature extracts it, yet rules walk through it untouched, for they reason from physics and owe the data nothing. So we build the bridge. Neuro-Symbolic Manifold Alignment (NSMA) embeds rule decisions as anchors inside the latent space of the neural policy, so that it keeps learning where learning pays, and can no longer forget what rules have always known. Generalization cannot be argued, only survived. We raise NSMA on 3G traces alone and release it, without fine-tuning, into eight unseen datasets spanning 4G, 5G, and WiFi, and onto a real-world player. It outperforms every state-of-the-art baseline. And when we open its latent space to ask why, probing and visualization return the same answer the design promised. https://tinyzqh.github.io/NSMA/ --- ### HPD-Parsing: Hierarchical Parallel Document Parsing Authors: Shu Wei, Jingjing Wu, Lingshu Zhang, Qunyi Xie, Hao Zou, Le Xiang, Xu Fan, Yangliu Xu, Manhui Lin, Xiaolong Ma, Cheng Cui, Tengyu Du et al. Published: 2026-07-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2607.18839v1 Efficient teamwork typically combines global coordination with parallel execution, a principle not yet fully reflected in unified Vision-Language Model (VLM)-based document parsers. Existing unified parsers process an entire page jointly but generate its output through a single token-by-token autoregressive trajectory, creating a sequential bottleneck that grows with document length. Such full-page sequential generation overlooks a key property of document parsing: layout must be analyzed globally, whereas block content can be parsed in parallel. Based on this observation, we introduce HPD-Parsing, which replaces full-page autoregressive generation with a Hierarchical Parallel Decoding paradigm. A main layout branch organizes the overall document structure and dynamically assigns block-level content decoding to concurrent branches, while progressive multi-token prediction (P-MTP) further reduces the decoding steps within each branch. Experiments on public benchmarks show that HPD-Parsing achieves 4,752 tokens per second, delivering $2.62\times$ the throughput of the fastest existing document parsing model and $3.06\times$ that of the vanilla autoregressive baseline, while maintaining competitive parsing accuracy. These results establish hierarchical parallel decoding as an effective alternative to full-page autoregressive generation, opening a new direction for efficient unified document parsing. --- ### ABOPD: Antibody CDR Design via On-Policy Distillation Authors: Zhuo Yang, Jiaying He, Jiaqing Xie, Daolang Wang, Xipeng Qiu, Yuxin Wang, Tianfan Fu, Beilun Wang Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18835v1 Antibodies are essential therapeutic molecules, and their complementarity-determining regions (CDRs) form the primary antigen-recognition interface. Recent protein generative models have demonstrated broad capabilities in biomolecular design, yet post-training strategies for downstream objectives remain limited. Standard denoising training operates on noisy states obtained by perturbing native structures, whereas recursive generation proceeds through model-generated intermediate states. For flexible antibody CDR loops such as CDR-H3, this mismatch can allow backbone deviations to accumulate along the denoising trajectory and compromise antigen-facing loop geometry. We introduce ABOPD, an antibody design framework based on on-policy distillation that leverages privileged native geometry during training to supervise states visited along the model's own denoising trajectories. With this fine-grained structural supervision, ABOPD substantially improves structural recovery on RAbD CDR-H3 generation, reducing RMSD by 0.42 Å (from 2.37 Å to 1.95 Å) and outperforming supervised fine-tuning and offline distillation controls, offering a path to higher-fidelity protein design. --- ### From Trajectories to Instructions: Language-Conditioned Meta-Reinforcement Learning Authors: Garvit Singla, Uma Maheswari Natarajan, Raghuram Bharadwaj Diddigi Published: 2026-07-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2607.18830v1 Model-Agnostic Meta-Learning (MAML) is a widely used framework for reinforcement learning (RL) that enables efficient transfer by learning global policy parameters that can be rapidly adapted to new tasks. MAML training proceeds in two loops: an inner loop where the global parameters are adapted to task-specific parameters, and an outer loop where these task-specific parameters are evaluated and losses are back-propagated to improve the global parameters. Traditionally, the inner loop adaptation is performed by collecting trajectories from the task environment and applying gradient updates on the empirical expected return, which can be a costly operation. We note that it is the outer loop that drives the actual learning of global parameters, and therefore the inner loop adaptation mechanism need not be restricted to be gradient-based. This observation leads us to ask: Can we replace the inner loop trajectory collection and gradient update with a simpler, task-specific signal? In many practical settings, tasks are naturally accompanied by language instructions. Leveraging these instructions as a direct task-specific signal, we propose LA-MAML (Language Adapted MAML), which modifies the inner loop by adapting the global policy parameters in a single step through a learned embedding of the task instruction, replacing the inner loop trajectory collection and gradient-based updates. Experiments on the BabyAI benchmark demonstrate that LA-MAML achieves competitive or improved performance compared to baselines at a significantly lower per-iteration wall-clock training time. These results demonstrate that language instructions are an effective and efficient substitute for trajectory-based inner loop adaptation in meta RL. --- ### Countercurrent Multiplier Networks: A Renal-Inspired Iterative Operator with Provably Bounded Fixed-Point Dynamics Authors: Snigdha Chandan Khilar Published: 2026-07-21 Categories: cs.LG, math-ph arXiv: https://arxiv.org/abs/2607.18829v1 The mammalian kidney concentrates urine using a mechanism with no analogue in current neural architectures: the countercurrent multiplier. Two anti-parallel flows joined at a hairpin recirculate a weak magnitude-bounded local pump into a large axial gradient achieving a four-fold concentration increase from a single-effect gradient that never exceeds 200 mOsm at any point. We formalize this mechanism as a differentiable sequence operator the Countercurrent Multiplier (CCM) layer and study it as an alternative to residual iterative refinement. --- ### Evaluating medical AI under missing information: same-provider judges and human raters change apparent safety Authors: Koyar Afrasyab Published: 2026-07-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.18828v1 Readiness stress-testing of medical AI has focused on closed-ended and multimodal benchmarks. We extend it to open-ended clinical conversation under missing information, where safe behavior means recognizing absent information and qualifying, clarifying, or not over-committing - and where the evaluator becomes part of the measurement. We stress-test four models - three flagships (Claude Opus 4.8, GPT-5.5, Grok 4.3) and one mid-tier model (Gemini 3.5 Flash) - by deleting the latter half of the final user turn in HealthBench conversations, grading responses with a four-provider LLM-judge panel and a blinded clinician-anchored reference. Two evaluator-facing results are robust. First, judge choice materially changes apparent safety: inter-judge agreement is only moderate (Fleiss' kappa = 0.65), and after adjusting for each judge's general leniency (vote-level logistic regression), a positive same-provider association remains (exact permutation p = 0.04; GPT-5.5 ~ +0.10 on the probability scale) - large enough to change which model appears to over-commit least once its own-provider judge is excluded. Second, LLM judges are more permissive than clinicians on a blinded 50-item subsample: all four are significantly more lenient than the stricter independent clinician (crediting appropriate uncertainty on 66-84% of items vs 52%), and three of four than the author-influenced consensus (Grok directional only; judge-vs-consensus kappa = 0.20-0.43). On the author-audited clinical-underdetermined subset the permissiveness gap widened and the point-estimate model ordering held. A closed-ended MedQA anchor confirms accuracy is high and option-order effects are within a +/-5-point equivalence region for three of four models, so the safety gap is about calibration, not knowledge. We release the harness, prompts, per-item outputs, judge panel, perturbation audit, and human-annotation protocol. ---