# Research Papers — 2026-10 · bundle page 30 of 156

> Full-text markdown of papers #1451–1500 of 7799 published in 2026-10, newest first. AI-readable mirrors of the full 7799-paper 2026-10 archive (50 papers per file).

Canonical page: https://aipod.dev/papers/archive/2026-10/page/30
[← Previous bundle](https://aipod.dev/papers/archive/2026-10/page/29.md)
[Next bundle →](https://aipod.dev/papers/archive/2026-10/page/31.md)

### Optimizing Team Behavior via Extremum-seeking Control in Multi-agent Reinforcement Learning

Authors: Svar Rajankumar Patel, Kshitij Jerath
Published: 2026-10-13
Categories: cs.MA, cs.AI, cs.RO, eess.SY
arXiv: https://arxiv.org/abs/2610.18808v1

In cooperative Multi-Agent Reinforcement Learning (MARL), agents must learn to coordinate their actions to achieve shared objectives. A key indicator of effective coordination is behavioral diversity, i.e., the agents' potential to enact distinct yet complementary control policies given the same set of observations. Although behavioral diversity can enable role specialization and richer joint behaviors, its appropriate level is highly context- and task-dependent and often costly to tune. We propose Adaptive Diversity Control (ADiCo), a two-timescale framework that adapts behavioral diversity online during training. On a fast timescale, Diversity Control (DiCo) regulates system diversity toward a target value. On a slower timescale, an Extremum Seeking Controller (ESC) adjusts this target using only mean episodic returns, without requiring an explicit model or analytical gradient of the diversity-performance landscape. We evaluate ADiCo on six configurations across two cooperative MARL tasks, where it maintains task-appropriate diversity levels: low diversity when homogeneous policies are beneficial and higher diversity when heterogeneous policies are required. Through this adaptation, ADiCo matches or exceeds the mean performance of DiCo (best grid-searched diversity target) and Unconstrained DiCo (diversity allowed to emerge without target regulation). These results demonstrate that incorporating classical feedback control into MARL training enables effective coordination by providing a systematic alternative to task-specific manual selection of the diversity target.

---

### SkillScaffold: Scaffolding Agentic Reinforcement Learning with Policy-Adaptive Skill Credit Assignment and Selective Injection

Authors: Wenlin Zhong, Hao Wang, Kun Yuan, Fan Yang, Jinluan Yang, Dian Jin, Xu Tan, Yiquan Wu, Kun Kuang
Published: 2026-10-13
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2610.18807v1

Agentic reinforcement learning (RL) enables large language models (LLMs) to learn from multi-turn interactions with verifiable outcomes. Textual skills can improve exploration by distilling prior interactions into reusable guidance. However, existing skill-based RL approaches often either assume skill injection is beneficial or rely on indirect feedback or rollout-intensive evaluation to assess its utility. Our pilot studies show that rollout-based utility estimation for skill credit assignment is costly and unstable under limited sampling, while skill benefits vary across candidates and depend on the current policy's competence on each query. To address these challenges, we introduce SkillScaffold, a policy-adaptive training scaffold that combines low-cost skill-utility estimation with selective skill injection. By reusing outcome-verified training trajectories, SkillScaffold generates candidate skills and estimates their utility through outcome-aware likelihood comparisons, without candidate-specific environment rollouts. The resulting utility guides whether and which skill to inject, retaining skill-free execution when no candidate is estimated to help. Joint training enables the Generator to produce skills tailored to the evolving Executor. Scaffolded RL encourages skill internalization for skill-free inference. Experiments on ALFWorld, WebShop, ScienceWorld, and $τ^2$-Bench demonstrate gains over strong agentic-RL and skill-based RL baselines across both short- and long-horizon tasks. Ablations and further analyses support the importance of utility-based selection and abstention. Our code is available at https://github.com/Vaidurya00/SkillScaffold.

---

### Registration is a Representation Problem: Solving Whole-Body PET/CT Registration

Authors: Neel Dey, Adrian V. Dalca
Published: 2026-10-13
Categories: cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2610.18804v1

Raw image intensities are a poor surrogate for similarity across images in correspondence problems. Further, while auxiliary segmentation labels can provide semantic identification and improve matching, they can assign only coarse identities to fine-grained anatomy and often cover only a subset of structures, particularly in whole-body settings. We present an approach that instead establishes correspondence using complementary representations of anatomy and demonstrate its utility for longitudinal whole-body PET/CT registration in the Learn2Reg 2026 PSMAReg Challenge. Our 3D foundation model, anatomix, is used to derive fine-grained, per-voxel semantic representations that are consistent across volumes, and we further employ off-the-shelf segmentation models to obtain coarse anatomical identities. Together, these representations drive a GPU-accelerated iterative registration solver (FireANTs) in a hybrid multistage framework: the first stage estimates alignment from dense multi-channel features, after which a second refines the deformation using both anatomix- and segmentation-based losses. Lastly, because registration should not introduce non-biological tumor volume changes across time, we develop lesion-specific component-wise volume preservation schemes. Without any task-specific training, our system won first place in the PSMAReg Challenge leaderboard. Code and interactive visualizations are available at http://www.neeldey.com/reps-for-reg.

---

### SEAR: Self-Evolving Agentic Robotics

Authors: Xinyi Wang, Bangzheng Li, Hanyu Wang, Lin Zhang, Andrea Yaoyun Cui, Zhenggang Tang, Shyamal Buch, Dejia Xu
Published: 2026-10-13
Categories: cs.RO, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.18792v1

Frontier language-model agents can now improve robots on their own: they write control code, test it, and keep what works as skills and memories for the next task. We introduce SEAR (Self-Evolving Agentic Robotics), a framework and benchmark built around two questions. Is the agent really learning, or just trying more often? And does what it learns still help on new tasks, in a new simulator, and on a real robot? The framework runs an agent in three modes: it writes code that becomes the policy (code-as-policy), acts live in the control loop (agent-as-policy), or works as a researcher that fine-tunes a vision-language-action model (auto-research). The benchmark adds controls that separate retained experience from repeated attempts, familiar tasks, privileged simulator state, larger budgets, and luck. Across seven frontier language models, 330 simulated and five physical tasks, and over four years of agent runtime, we highlight three findings. First, the learning curve of a robotic agent depends both on the overall time budget and the task difficulty relative to model capability. Harder tasks make longer sessions worthwhile; on easier tasks, re-rolling short attempts does just as well. Second, control code and shared routines transfer across tasks and simulators, while task-specific solutions do not. On a real robot, we observe a performance gap between agent-designed practice and novel human-designed scenes. Third, different models evolve more effectively in different ways. Some do best acting live, others writing code first. And a text-only model keeps pace with a comparable multimodal one: perception helps with error assessment, but agentic reasoning and tool use can matter more. We release the benchmark and codebase to support future research.

---

### Trust Before Autonomy: Rethinking Agentic Remote Sensing for Crisis Response

Authors: Georgios Angelidis, Aggelos Psiris, Spyridon Evangelatos, Thomas Lagkas, Panagiotis Sarigiannidis, Iraklis Varlamis, Georgios Th. Papadopoulos
Published: 2026-10-13
Categories: cs.CV, cs.CY
arXiv: https://arxiv.org/abs/2610.18789v1

Agentic remote sensing is evaluated on how many questions it can answer, whereas crisis response depends on whether its answers can be relied on and how much of the decision can safely be handed to the agent. This paper argues that \emph{autonomy should be earned, not assumed}: An Earth-Observation (EO) agent should be granted autonomy in proportion to demonstrated reliability, epistemic honesty and verifiability. The position is supported by an evidential meta-analysis of twelve agentic RS/EO/GIS benchmark studies from peer-reviewed venues. Only one repeats runs, three contain unanswerable tasks and none measures calibration. Their own tables show that completed workflows are often not correct (task success exceeds tool-call correctness by up to 54pp on identical queries), that the gap to an expert trajectory widens overall with trajectory length (from 7 to 56pp), and that capability does not predict abstention (Spearman r from -0.25 to 0.50, all n.s.). The paper then contributes three EO-specific trust properties, an autonomy ladder for crisis EO tasks, and a readiness protocol that replays archived emergency activations before any live use. In this context, current evidence justifies assistive roles only.

---

### Bytes to Threats: Byte-Sequence Transformer For Malware Detection

Authors: Abdelrahman Yassin Eljamal, Abdeltawab Hendawi
Published: 2026-10-13
Categories: cs.CR, cs.AI
arXiv: https://arxiv.org/abs/2610.18787v1

Existing malware detection approaches heavily rely on handcrafted features derived from static or dynamic analysis of the executable files. These methods require specialized domain expertise, dedicated analysis tools, and time-consuming analysis, which limits their scalability and robustness in the face of the increasing complexity and diversity of modern malware. Consequently, there is a growing interest in approaches that operate directly on raw bytes, which contain rich semantic and structural information. However, capturing long-range dependencies within executable files remains challenging for conventional static and CNN-based methods. This paper presents MalwareFormer, a transformer-based, byte-level malware detection framework that operates directly on raw bytes, without requiring domain knowledge or feature extraction. MalwareFormer achieves an accuracy of 98.35%, a balanced accuracy of 98.32%, a macro-F1 of 98.32%, and a false positive rate of 1.94%, outperforming both baselines across all threshold-dependent classification metrics and demonstrating that transformers provide a powerful and scalable approach for byte-level malware detection.

---

### Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation

Authors: Donghwan Kim
Published: 2026-10-13
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2610.18786v2

LLM ensemble gains are often measured against a single baseline, but the value attributed to additional candidates can change with the comparison used. We study an LLM aggregator under four matched conditions: Direct (no candidates), Single-slot (one candidate in a multi-slot interface), Duplicate (one candidate repeated), and Multi-candidate (separately generated candidates). Under either non-diverse control, overall candidate gain (CIV) splits algebraically into an additional-candidate contrast (DSV) and a candidate-context contrast (CCV); the split, but not CIV, depends on the chosen control. On MultiHopRAG model diversity (k=2), the repetition gap Rep=-8.33 pp (-13.25, -3.75) quantifies the difference between gains measured against Duplicate and Single-slot. On a matched FRAMES stronger-panel extension (n=200, k=3, aggregator output limit 4096), paired Delta Rep=+9.67 pp (+6.00, +13.50); within panels, Rep changes sign (original -6.33; stronger +3.33). In FRAMES evidence aggregation, candidate-context contrasts remain strongly negative under both non-diverse controls. Our results show why ensemble evaluations should state the reference control before attributing gains to additional candidates.

---

### Where Do LLM Ensemble Gains Come From? Decomposing Candidate Value, Diversity, and Aggregation

Authors: Donghwan Kim
Published: 2026-10-13
Categories: cs.CL, cs.AI
arXiv: https://arxiv.org/abs/2610.18786v1

LLM ensemble gains are often measured against a single baseline, but the value attributed to additional candidates can change with the comparison used. We study an LLM aggregator under four matched conditions: Direct (no candidates), Single-slot (one candidate in a multi-slot interface), Duplicate (one candidate repeated), and Multi-candidate (separately generated candidates). Under either non-diverse control, overall candidate gain (CIV) splits algebraically into an additional-candidate contrast (DSV) and a candidate-context contrast (CCV); the split, but not CIV, depends on the chosen control. On MultiHopRAG model diversity (k=2), the repetition gap Rep=-8.33 pp (-13.25, -3.75) quantifies the difference between gains measured against Duplicate and Single-slot. On a matched FRAMES stronger-panel extension (n=200, k=3, aggregator output limit 4096), paired Delta Rep=+9.67 pp (+6.00, +13.50); within panels, Rep changes sign (original -6.33; stronger +3.33). In FRAMES evidence aggregation, candidate-context contrasts remain strongly negative under both non-diverse controls. Our results show why ensemble evaluations should state the reference control before attributing gains to additional candidates.

---

### Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents

Authors: Yuchen Lu, Julian McAuley, Lei Zhang
Published: 2026-10-13
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.18784v2

Long-horizon agents accumulate progress in persistent task state, yet repeatedly process interaction histories describing how that progress was achieved. This raises a natural question: can long-horizon context instead be organized around reusable progress rather than the history that produced it? In this paper, we introduce CarryOn, an object-oriented context runtime that organizes continuation around reusable task objects. Rather than treating compressed histories or reconstructed accounts of progress as the primary basis for continuation, CarryOn builds a context view from continuation carriers, complementary execution context, and a live tail of recent interactions. A core challenge is keeping this reusable basis valid after its surrounding history is retired: how can CarryOn update the basis without reconstructing the full representation at every fold? CarryOn maintains carrier records across folds, revising affected records against bound evidence and applicable environment checks while carrying forward unaffected records and jointly updating execution context. Experiments on OSWorld 2.0 and BrowseComp-Plus show improved gains of 5.8% and 7.0% while reducing total token usage by 51.9% and 22.9%, respectively. Project page: https://ethannju.github.io/carryon

---

### Pack Light, CarryOn: Building Object-Oriented Context for Long-Horizon Agents

Authors: Yuchen Lu, Julian McAuley, Lei Zhang
Published: 2026-10-13
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.18784v1

Long-horizon agents accumulate progress in persistent task state, yet repeatedly process interaction histories describing how that progress was achieved. This raises a natural question: can long-horizon context instead be organized around reusable progress rather than the history that produced it? In this paper, we introduce CarryOn, an object-oriented context runtime that organizes continuation around reusable task objects. Rather than treating compressed histories or reconstructed accounts of progress as the primary basis for continuation, CarryOn builds a context view from continuation carriers, complementary execution context, and a live tail of recent interactions. A core challenge is keeping this reusable basis valid after its surrounding history is retired: how can CarryOn update the basis without reconstructing the full representation at every fold? CarryOn maintains carrier records across folds, revising affected records against bound evidence and applicable environment checks while carrying forward unaffected records and jointly updating execution context. Experiments on OSWorld 2.0 and BrowseComp-Plus show improved gains of 5.8% and 7.0% while reducing total token usage by 51.9% and 22.9%, respectively. Project page: https://ethannju.github.io/carryon

---

### Designing for Prediction: Planning Formulas for Expected Out-of-Sample $R^2$ and $r$

Authors: Elizabeth Tipton
Published: 2026-10-13
Categories: stat.ME, stat.ML
arXiv: https://arxiv.org/abs/2610.18781v1

Design calculations let researchers gauge, before data is collected, whether a planned study is likely to detect an effect. In this paper, we develop such calculations for predictive models using ordinary least squares and ridge regression under leave-one-out cross-validation. To do so, we derive closed-form expressions relating the expected out-of-sample $R^2$ and the expected correlation between observed and predicted outcomes to four design-stage quantities: sample size, number of predictors, an assumed population $R^2$ ($ρ^2_{in}$) and a design efficiency factor $D^*$ (which summarizes how leverage is distributed across observations). Unlike the classical shrinkage formulas, which depend only on sample size and dimension, these are sensitive to how covariates are arranged -- to imbalance and skew. A simulation study assesses the accuracy of the formulas against realized values under a variety of data generating mechanisms. We apply the methods to a published study predicting personality from smartphone behavior and show that the design formulas would have indicated, before any analyses were run, that least squares could not be expected to beat the sample mean while ridge could. An R package, predpower, implements the calculations.

---

### BLADE: Auditing Information Use for Failure Attribution in LLM Multi-Agent Systems

Authors: Xiaoguang Guo, Yaohao Long, Chuxu Zhang
Published: 2026-10-13
Categories: cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.18780v1

Large language model multi-agent systems extend individual agents' capabilities through collaboration across specialized roles, tools, and workflows. Yet collaboration makes failures harder to diagnose: one agent may misinterpret a correct result while others carry the mistake forward. Failure attribution therefore seeks to identify the responsible agent and decisive error step. Recent methods trace dependencies, but knowing where information came from is not the same as knowing where it went wrong. We introduce BLADE (Blame Localization via Audited Dependency Evidence), a training-free framework that moves beyond tracing information to auditing its use. Using Jev for structured probabilistic judgments, BLADE uses dependency probabilities to select relevant source-decision handoffs for auditing. At these handoffs, it examines whether receiving decisions use their sources faithfully and within task requirements, distinguishing error introduction, propagation, and repair. These findings inform a global diagnosis that considers all execution steps. Without reference answers, BLADE achieves the best or tied-best step accuracy among eight strategies under a shared Jev decision interface on 1,228 failed runs across five Who&When and Who&When Pro subsets, outperforming All-at-once by 2.2-11.1 percentage points. Making responsibility evidence explicit offers a path toward more interpretable and reliable agentic systems.

---

### GAD: Geometry Attention Distillation for Consistent Video Generation

Authors: Junfeng Ni, Ziyu Zhu, Fan Jiang, Yu Liu, Chiyu Wang, Jiawei Zhang, Zhaoxu Sun, Mu Xu, Siyuan Huang, Yixin Chen
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18779v1

Video diffusion models generate visually compelling short-term sequences but often struggle to preserve scene geometry and object structure over time. In this paper, we first demonstrate that current video attention cannot identify reliable correspondences across frames, with responses dispersed across multiple regions even after recognizable scene structure has emerged. We argue that reliable cross-frame associations are essential for information exchange as scene structure forms in video generation. Thus, we introduce Geometry Attention Distillation (GAD), which transfers correspondences from a frozen geometry model's attention to video attention. Extracted from clean videos, these correspondences link the same static surface or moving object across frames. Since structure formation begins in deeper layers at higher noise and shallower layers at lower noise, we apply attention distillation at a student layer selected according to the noise timestep. After training, video attention focuses more consistently on corresponding locations. Experiments across all three video backbones demonstrate improved long-range scene consistency and dynamic temporal consistency, with geometric gains across generation lengths. On Wan2.2 5B, GAD improves cross-frame matching accuracy by at least 30% over the Wan baseline on both long-range benchmarks. Notably, GAD achieves these gains without additional inference modules, offering a promising approach to efficient and consistent video-based world simulation. Project page: https://fantasy-amap.github.io/GAD/.

---

### Jeva: System One Models from First Principles

Authors: Qi Li, Xinchao Wang
Published: 2026-10-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.18773v1

System one models enable direct, low-latency decision-making. Yet recent practice of predicting an answer from serialized text does not, by itself, respect the structure of a decision problem. We argue that current designs underemphasize four important structural requirements. (i) null-aware: Models should distinguish answer absence from uncertainty among candidates; (ii) permutation-invariant: Reordering alternatives should not change the decision; (iii) option-scalable: Growing candidate sets should not require redefining the output space; (iv) cacheable: Request-independent computation over recurring alternatives should be reusable across requests. We introduce Jeva, a decision-native architecture that addresses these requirements jointly. A shared encoder independently represents the request and each option, while a compact decision stack refines only the request state against immutable option memory. Symmetric aggregation yields permutation-equivariant probabilities; shared scoring accommodates variable-size candidate sets. A separate coverage estimate distinguishes answer absence from candidate competition, enabling explicit null outcomes. Request-independent option representations and layer projections support reuse across newly assembled sets. These mechanisms unify Choice, Noul, and Score outputs. Across evaluations, Jeva combines strong typed-decision performance with near-perfect null detection and no decision flips under any tested permutations. It scales to thousands of candidates and substantially reduces latency through cross-request reuse without changing decisions.

---

### The Federated Tumor Segmentation and Classification (FeTS++) Challenge

Authors: Wenyi Tang, Carla Pitarch-Abaigar, Siddhesh Thakur, Verena Chung, Konstantinos Koukoutegos, Alexandros Karargyris, Holger R. Roth, Spyridon Bakas
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18771v1

International challenges have become the standard for developing and benchmarking biomedical image analysis methods using centralized multi-institutional data. Federated learning (FL) describes a paradigm-shift on multi-institutional collaborations, where models are trained across institutions while their data are always retained locally. The Federated Tumor Segmentation (FeTS) challenge, first held in 2021, was the first challenge dedicated to FL in medical imaging and editions in 2022 and 2024 pushed the frontiers of real-world method validation. FeTS++ 2027 returns after a two-year gap with a substantially larger and more diverse cohort, and with the two "pluses" that name the challenge: an official classification task alongside segmentation, and a move of the infrastructure to NVIDIA FLARE (NVFlare). Built on the clinically acquired, multi-institutional multi-parametric MRI data of the BraTS 2025 Lighthouse challenge, FeTS++ spans gliomas and meningiomas across patient cohorts from institutions in North America, Europe, Asia, and Africa, and focuses on ranking submitted methodological approaches according to their model utility and an efficiency-aware score. Task 1 asks participants to find the best strategy for aggregating knowledge from per-institution segmentation models, while Task 2 asks them to distinguish glioma from meningioma under the same federated protocol, in which the entity label is heavily confounded with the site of origin.

---

### MIST: Multi-Scale Spatial Context for Gene Expression Prediction from Histology

Authors: Koushik Howlader, Tirtho Roy, Md Tauhidul Islam, Wei Le
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18769v1

Predicting spatial gene expression from standard H&E images could extend molecular analysis to tissues without spatial transcriptomic measurements. A standard Transformer applies self-attention over spots but ignores where each spot sits in the tissue, whereas a spot's molecular state depends on tissue context at multiple spatial scales. Building on this insight, we introduce MIST, a Multi-context Spatial Transformer that augments the self-attention backbone with two context pathways applied in every layer: a coordinate-invariant local pathway that attends over spatial k-nearest-neighbor spots, and a slide pathway that broadcasts a whole-slide summary to every spot; the model predicts all spots in a single forward pass. Because coordinates enter only through pairwise distances, the local pathway is invariant to translations, rotations, and reflections when image features are fixed. In intra-cohort evaluation across six human organ cohorts from HEST-1k, MIST improves mean Pearson correlation over the strongest evaluated baseline by 9.7%, 11.6%, and 11.4% for 10-, 50-, and 100-gene panels, respectively, with the lowest average MSE and MAE. MIST further improves leave-one-organ-out transfer by 6.0% over the next strongest baseline, attains the best expression-derived domain recovery (highest mean ARI and NMI at every clustering resolution, and higher NMI than the closest baseline on 84% of held-out slides), and reproduces spatial expression patterns of selected cancer-marker genes. Together, these results show that adding local and slide context to a vanilla Transformer backbone is an effective approach to spatial gene expression prediction from histology.

---

### Rubrics from the Task, Verdicts from Evidence: Pointwise Rubric Rewards with Tool-Augmented Judges

Authors: Tianci Liu, Haoyu Wang
Published: 2026-10-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.18766v1

Rubric-based reward models (RMs) evaluate responses against explicit criteria. However, existing methods often derive criteria from observed response candidates, introducing response-induced bias that may limit rubric generalizability. Their judges also struggle with mechanical constraints when relying on reading alone, and commonly used pairwise judges provide only relative preferences, complicating their direct use for policy optimization. We propose ReVeR, a pointwise rubric RM that derives hard rules and graded principles by reasoning about the instruction alone and evaluates responses independently through tool-augmented judging. Both components are trained through label-blind distillation, in which preference labels only filter a teacher's outputs, yielding ReVeR-Traces, a large-scale dataset of rubric-reasoning and judging trajectories. Experiments demonstrate that ReVeR outperforms pairwise rubric RMs in average preference accuracy, most clearly on instruction following, as well as in test-time scaling and downstream policy optimization. We further find that generator RL improves preference accuracy while reducing rubric fidelity, and that higher preference accuracy does not necessarily translate into better downstream policies. These findings highlight that rubric fidelity, preference accuracy, and downstream utility are distinct properties, shedding light on future RM research.

---

### LayerCanvas: Infographic Generation in AI Learning Assistants via Hybrid Layer-Set Diffusion

Authors: Keshav Gupta, Jeremy Owen, Christopher Lloyd, Shaunak Mishra, Manav Sharma
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18763v1

Online learning platforms are increasingly deploying AI assistants to provide personalized, on-demand tutoring. These can leverage infographics to communicate complex information and improve the learning experience. However, automating effective infographic generation remains a challenging task, requiring a combination of visual richness with precise typography, structured layouts, and continued editability. Existing representation methods satisfy these goals only in part: Scalable Vector Graphics (SVG) representations preserve geometry, hierarchy, and text but provide limited aesthetic flexibility, while raster diffusion models produce richer appearances by flattening content and layout into a single image. Conditioning a diffusion model on a rasterized SVG does not resolve this conflict because the raster preserves approximate placement but discards element ownership, occlusion relationships, and the structure needed to modify and recompose the result. We introduce LayerCanvas, a hybrid approach that retains SVG as the semantic, structural, and typographic program while using diffusion to aestheticize non-text content. Rather than generating one flattened raster, LayerCanvas partitions SVG elements into z-ordered sets of mutually non-overlapping elements and processes them through communicating parallel RGB generation streams. The resulting layers are independently editable and deterministically recomposed, while text is rendered from preserved vector metadata. Our approach combines rich visual synthesis with accurate typography and localized editing. LayerCanvas substantially improves text fidelity over state-of-the-art raster image generators while achieving the highest overall aesthetic quality. We further demonstrate non-destructive content editing, localization, asset reuse, and animation.

---

### Looking Complete Is Not Being Grounded: Evaluating LLM Sitreps Against Human-Authored Situation Reports

Authors: Sachin Kumar
Published: 2026-10-13
Categories: cs.CL, cs.CY, cs.SI
arXiv: https://arxiv.org/abs/2610.18762v1

Evaluation for automatic crisis reporting should measure what practitioners need, not what is convenient to compare. Through LLM-assisted, author-approved content analysis, we derive a 15-slot information schema from professional humanitarian situation reports using qualitative content analysis, then evaluate 121 human and 261 LLM-generated reports across five disasters and three hazard types. Under google/gemini-3.7-flash on Cyclone Idai, the best machine arm is nominally above the empirical human ceiling (0.827 vs. 0.821), but the difference is statistically indistinguishable from zero (95\% bootstrap CI $(-.049,.059)$; permutation $p=.852$), and the ordering does not replicate under openai/gpt-5.6-luna. On a blind 100-item Idai sample, the author agrees with the Gemini judge at 0.910 accuracy and unweighted Cohen's $κ=.799$. All four judges clear the plan's raw-agreement criterion of .7; only Gemini reaches .7 on chance-corrected $κ$. Across the three primary events, machine reports reproduce 5.4-11.0% of same-day official figures. Across those same events, raw exact-number matching links 14.3-19.6% of official figures and 85.4-90.1% of machine figures to the social stream; manual audit confirms 45/50 declared absences, 4/25 official matches, and 24/25 machine matches semantically. The issuing-provenance gap is positive in all 12 observed event-judge cells: report date, organisation, series, and authority were not supplied at professional-report coverage in any cell. On the three primary events, ROUGE-L is uninformative, BERTScore is negatively associated with completeness after multiplicity correction, and a generic LLM judge changes sign.

---

### Beyond the Verdict: Agentic Code Judging with Verifiable Feedback

Authors: Zheye Deng, Sha Li, Changlong Yu, Xin Liu, Qin Lu, Priyanka Nigam, Yangqiu Song
Published: 2026-10-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.18760v1

Test-time scaling for large language models (LLMs) can involve generating multiple candidate solutions for selection or revising an earlier attempt. To guide selection and revision, we need a verifier that can estimate candidate correctness and provide evidence of failure. In competitive programming, however, official judging systems provide authoritative verdicts under finite submission budgets, without revealing failure explanations or failing inputs. To provide informative feedback without consuming official submissions, we formulate a task in which a model judges and diagnoses a candidate program using only the problem statement and source code, with generated counterexamples serving as checkable evidence. To evaluate this task, we build USACO-JUDGE from the USACO 2026 contests, using accepted and unaccepted programs written by human experts and multiple LLMs. We then train ARBITER, a 35B mixture-of-experts (MoE) agentic verifier, with reinforcement learning in a generic execution sandbox. Training raises counterexample success on incorrect programs from 40.61% to 62.67% and mean binary verdict accuracy on USACO-JUDGE from 71.41% to 87.41%, approaching frontier verification performance. Beyond verification, ARBITER outperforms baselines in both code selection and refinement. In an IOI 2026 case study, ARBITER-guided refinement raises GPT-OSS-120B's expected score above the gold-medal threshold under the official submission limit.

---

### PreciseCoder: Faithful and Generalizable Debugging Beyond Unit-Test Rewards

Authors: Yunkai Zhan, Miaosen Chai, Shangshang Wang, Jike Zhong, Yuqing Yang, Deqing Fu, Serina Chang, Wang Bill Zhu
Published: 2026-10-13
Categories: cs.SE, cs.AI, cs.CL
arXiv: https://arxiv.org/abs/2610.18755v1

Reinforcement learning (RL) from unit tests is the common recipe for post-training code models, but a unit test only asks whether the final program passes. We analyze what this unit-test-only reward teaches a model during debugging and find that, while it raises pass rates, it also teaches the model to edit more and to hallucinate bugs in code that is already correct. We introduce PreciseCoder, a training algorithm that augments the unit-test reward with a signal for semantically verified, localized edits, which correlates with smaller edits and higher pass rates. On 4B, 9B, and 27B models, PreciseCoder reduces the average number of edit lines from 7-15to 3 at a pass rate on par with unit-test-only training. Trained only on single-line bugs, PreciseCoder can generalize to repository-level program repair, where it resolves more issues with smaller patches. Its precision also transfers from weaker to stronger models: given the reasoning trace of a 9B PreciseCoder model, frontier models such as GPT-5.6 Sol and Qwen3.8-Max gain 4.6-12.4 points in edit precision with no additional training. Linguistic analysis of the thinking traces shows that our reward encourages the model to reason about individual lines, name the buggy ones, and form consistent bug hypotheses.

---

### SolBehave: Evaluating the Behavioural Correctness of LLM-Generated Smart Contracts

Authors: Giovanni Pio Delvecchio, Stefano Fantazzini, Giacomo Frisoni, Gianluca Moro, Lorenzo Bacchiani, Mario Bravetti, Thomas Troels Hildebrandt
Published: 2026-10-13
Categories: cs.SE, cs.AI
arXiv: https://arxiv.org/abs/2610.18745v1

Large language models increasingly generate Solidity smart contracts that, once deployed, can never be patched, making a single latent fault permanent and exploitable. Yet current evaluations mostly ask only whether generated code compiles. We introduce SolBehave, a benchmark for evaluating the correctness of model-generated contracts at scale, moving beyond ad hoc unit tests. Across 250 real, reproducible Ethereum contracts, we apply three complementary tests of behavioural fidelity: differential fuzzing, historical-transaction replay, and, as our core contribution, synthetic-transaction replay guided by a Dynamic Condition Response (DCR) graph that enumerates the space of legitimate function-call sequences. Differential fuzzing compares each generated function with the original on automatically generated inputs; historical replay compares the whole contract on the transactions it actually received on-chain; DCR-guided replay compares it on sequences of calls that follow its intended usage, including orderings that never occurred on-chain. We evaluate seven state-of-the-art open-weight models after one round of automated repair. Compilation success ranges from 26.4% to 92.7%, but under differential fuzzing only 31.4% of all generated function bodies that compile, pooled over models and samples, are equivalent to the original. Divergence at the contract level is the norm rather than the exception: for six of the seven models, between 13% and 31% of the generated contracts that compile diverge from the original under historical replay, and between 13% and 44% under DCR-guided replay; only Gemma-4-31B stays below 3% on both. Compilation is therefore a necessary condition for correctness and far from a sufficient one.

---

### Emergent Loop Geometry in Gauge-Equivariant Graph Neural Networks

Authors: Ali Rayat, Gia-Wei Chern
Published: 2026-10-13
Categories: cond-mat.stat-mech, cond-mat.str-el, cs.LG, hep-lat
arXiv: https://arxiv.org/abs/2610.18742v1

Gauge-equivariant graph neural networks generate ordered products of gauge links through local message passing, but the geometry of these internal structures is not captured by the receptive field alone. We investigate the statistical ensemble of closed lattice walks generated by recursive node- and edge-feature operations, independently of network training. Analytical and numerical results reveal a separation between rapid growth of contour length and much slower spatial expansion: deeper compositions generate increasingly long paths through repeated traversal of relatively compact regions. Loop expansions of fermionic observables provide a physical interpretation of this hierarchy, showing how spatial locality remains compatible with high-order dependence on nearby gauge fields. These findings identify complementary roles of network depth in extending spatial reach and enriching the representations constructed within that reach, providing a geometrical perspective on learning lattice gauge observables.

---

### Context-aware Bayesian Fusion for Multimodal Object Detection

Authors: Arunachalam Venkatachalam, Mhafuzul Islam, Marwan Yusuf, David Van Hamme, Ben Stoffelen, James Foresi
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18741v1

Reliable detection of pedestrians and road users is essential for the safe operation of autonomous vehicles. However, this task becomes increasingly challenging under adverse environmental conditions, such as poor weather and low lighting, where the performance of state-of-the-art pre-trained object detectors and on-board sensors tends to degrade, which limits their effectiveness and reliability. Moreover, retraining models for each operating context imposes significant constraints due to high data requirements, computational resources, and time. Considering this, we propose a novel context-aware multimodal late fusion approach that integrates a lightweight gating network within our probabilistic calibration framework. The gating network extracts environmental context and dynamically modulates the detection outputs of pre-trained object detectors based on the inferred context, without retraining of the detectors. Our methodology is evaluated on a modified KITTI dataset, in which the camera images and lidar pointclouds are augmented for varying fog levels. Experimental results show that our approach yields an approximately 14% relative improvement over non-adaptive late fusion baselines, achieving a consistent average precision of 56% across different fog levels.

---

### A Rate Transfer Method: Convergence Rates in Stochastic Gradient Descent via Stochastic Calculus

Authors: Neil Walton
Published: 2026-10-13
Categories: cs.LG, cs.AI, math.OC, math.PR, stat.ML
arXiv: https://arxiv.org/abs/2610.18739v1

We develop a rate-transfer method that converts finite-time Lyapunov estimates for an associated stochastic differential equation (SDE) into finite-time convergence bounds for stochastic gradient algorithms. Under generic moment and smoothness conditions, the convergence analysis of the stochastic gradient method reduces to a calculus argument verifying a Lyapunov generator inequality for the SDE. The method recovers established rates for classical stochastic gradient methods and provides a way to establish convergence rates for new optimization algorithms through calculus arguments applied to their associated SDEs.

---

### Enabling Token-wise Importance Sampling in Off-Policy RLVR with Proximal KL Regularization

Authors: Hao Bai, Jiarui Yao, Tong Zhang
Published: 2026-10-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.18738v1

Off-policy data in reinforcement learning with verifiable rewards (RLVR) can arise from asynchronous generation, rollout-training mismatch, and response reuse. Token-wise importance weighting avoids multiplying likelihood ratios across a response but generally biases policy gradients. We prove that, under coverage and local expressiveness conditions, full future KL credit makes the exact proximal optimizer the update's unique stationary policy, even for mixtures of older policies, without full-response weights or a fitted critic. We further establish convergence of continuous-time tabular population updates under persistent coverage and time-varying behavior policies, and a finite-sample reward guarantee for linear logistic models under additional assumptions. These results motivate Proximal KL, a practical off-policy RLVR method that separates the proximal reference from the sampling policy and replaces PPO clipping with bounded future KL credit. Across all seven matched off-policy comparisons on Qwen3-1.7B and Qwen3-4B, spanning two learning rates, response reuse, and rollout staleness, Proximal KL outperforms GRPO in micro-averaged accuracy over eight math benchmarks by up to 2.58%. A separate on-policy comparison yields comparable aggregate accuracy. In periodic-refresh experiments, Proximal KL avoids the collapses observed in PG and CISPO, exceeds GRPO's training accuracy, and reaches higher accuracy in fewer policy updates. Ablations show that token-wise correction outperforms sequence-level and unweighted controls. With token-wise correction retained, future KL feedback reduces sensitivity to the tested baseline perturbation relative to PPO clipping. A finite future horizon improves AIME24 mean@16 and pass@16 over the full tail, while fixed-policy measurements show that truncation reduces correction variance over the tested horizons.

---

### Concealed Handgun Detection using WiFi Signals

Authors: Erik Persson, David Gustafsson, Tomas Wilkinson
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18736v1

While conventional camera-based surveillance systems and metal detectors can provide a sufficient level of security when used appropriately, they have fundamental limitations that reduce their applicability. Camera-based systems are unable to detect occluded or concealed threats and metal detectors require each person to enter through a narrow opening one at a time, leading to congestion issues and limited use in large crowds. A WiFi-based security system has the potential to be a complementary modality that is able to handle occlusions or concealment and is not hindered by strict spatial configurations. In this paper, we propose use Channel State Information estimated from WiFi-signals to detect concealed handguns. To this end we collect a dataset of aligned WiFi signals and RGB images that we call the WiFire dataset. We also propose a deep learning model called WiFireNet tailored to detect concealed handguns in WiFi signals using vision-based supervision signals. Our results show that concealed handgun detection is possible in WiFi and has promise as a complementary modality to surveillance cameras and metal detectors. We release the dataset, model and code to facilitate the further research into WiFi-based handgun detection.

---

### SP$^2$ec: Adaptive Self-Speculative Decoding for Vision-Language Models

Authors: Yuqi Huang, Xingyao Li, Yunlong Hou, Fengzhuo Zhang, Jiachun Pan, Vincent Y. F. Tan
Published: 2026-10-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.18735v1

Speculative Decoding (SD) has demonstrated its efficacy in accelerating large language model inference, but for Vision-Language Models (VLMs), the diverse input modalities make it difficult for a small trained drafter to achieve consistently high throughput. Self-speculative decoding reuses parts of the target model as a drafter which retains the multi-modal capability of the VLMs, yet its efficiency depends on choosing a suitable draft configuration for each prompt. We propose Single-Peak SPeculative decoding (SP$^2$ec), a training-free bandit algorithm that adapts the draft configuration during generation. SP$^2$ec exploits the single-peak throughput pattern observed in existing self-speculative methods, including SWIFT, KnapSpec, SpecVLM, and our simple cosSim. Our algorithmic design explicitly minimizes a novel notion of wall-time regret for each given query, corresponding to the empirical wall-time latency. Theoretical guarantees on the wall-time regret show that SP$^2$ec can exploit the optimal draft configuration, with negligible exploration cost. Extensive experiments on Qwen2.5-VL-7B-Instruct and LLaVA-NeXT-Mistral-7B across visual and textual tasks show consistent improvements of up to 11\% over the original self-speculative methods, demonstrating the wide applicability of the proposed SP$^2$ec algorithm.

---

### From Fourier Amplitudes to Persistence Images for Impact-Echo Defect Detection

Authors: Keunsu Kim, Suyoung Choi, Meiyan Kang
Published: 2026-10-13
Categories: stat.ME, stat.AP, stat.ML
arXiv: https://arxiv.org/abs/2610.18732v1

Impact-echo testing is a nondestructive testing method used to detect internal defects in concrete structures. However, boundary reflections, structured interference, and measurement noise can produce complex Fourier amplitude spectra, making defect classification difficult. We constructed a two-dimensional frequency--amplitude persistence image (FA-PI) from the Fourier amplitude spectrum. Using a fixed parameterization of exact multi-parameter persistent homology, each retained Fourier component is represented with frequency-bin location as one coordinate and amplitude as the other, and Gaussian kernels are then applied in this two-dimensional space. This construction has a direct analytical connection to the corresponding one-dimensional Gaussian-smoothed Fourier amplitude spectrum. FA-PI was evaluated on an experimental backfill-void dataset and a public bridge-deck impact-echo dataset, and compared with raw and Gaussian-smoothed Fourier amplitude spectra using a linear classifier with a Softmax output and identical signal-level data splits. The evaluation included original signals and signals corrupted by Gaussian noise, sinusoidal interference, or mixed noise, with matching noise types and signal-to-noise ratios during training, validation, and testing. Performance was assessed using accuracy and macro-averaged precision, recall, and F1-score. Across the two datasets, FA-PI consistently outperformed the raw and Gaussian-smoothed Fourier baselines in the noisy-condition summaries. These findings indicate that representing Fourier spectral components in a two-dimensional frequency--amplitude space can improve impact-echo defect classification beyond frequency-axis smoothing alone.

---

### Learning When to Transfer: Reinforcement Learning for Carbon-Footprint Modeling in New Manufacturing Domains

Authors: Muhammad Umar Farooq, Daniel R. Cooper
Published: 2026-10-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.18730v1

Manufacturers often need to characterize a new material, process, or product before enough operating data are available to train reliable predictive models for this new, target domain. Existing manufacturing operations contain potentially useful knowledge to accelerate this characterization, but indiscriminate reuse of data from existing source domains can cause negative transfer in learning of the new, target domain when the manufacturing domains differ sufficiently. This work proposes a reinforcement-learning method for selective knowledge transfer, where its reward scores each opportunity to transfer against supervised alternatives, so negative transfer is avoided and controlled during forward prediction instead of being diagnosed afterwards. Given limited data from a new target domain, the learned policy chooses among predictors based on individual source operations, subsets of the sources, or the supervised alternatives: fully pooled sources, or target data alone. A validation gate deploys the transferred prediction only when it improves upon the supervised alternatives. The method is designed for learning variables supporting carbon accounting, as well as performance and quality outcomes. We demonstrate the approach by quantifying the embodied greenhouse-gas emissions of parts produced via material-extrusion additive manufacturing, using data from 729 experiments spanning five materials: ABS, PETG, TPU, PLA, and recycled PLA. The response combines measured fabrication electricity with polymer-specific cradle-to-gate production factors, while 22 material, process, and toolpath features characterize each experiment. In leave-one-material-out evaluation, this method achieves a mean test RMSE of 5.2 g CO2e per part, 20.00% lower than target-only training (32.5% lower when only 5-10% of the target training data are available) and 10.30% lower than indiscriminate source-target pooling.

---

### Segment Where You Look: Attention as the Interface Between Language Models and Segmenters

Authors: Eslam Abdelrahman, Hamza Omar Alkaf, Yuheng Li, Siddharth Srinivasan Iyer, Nanxuan Zhao, Eli Shechtman, Krishna Kumar Singh, Mohamed Elhoseiny
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18729v1

Vision-language models (VLMs) that segment from language must pass the location of the referent to a pixel decoder. The dominant interface compresses it into the hidden state of a single (SEG) token, a d-dimensional vector with no spatial index, and leaves the decoder, which never sees the expression, to recover the referent's extent. We show that the VLM already computes a better signal: the (SEG) token's attention over the image tokens, which is defined on the patch grid by construction. Probing a frozen segmentation VLM, we find that this attention yields 18 to 22 more gIoU than the (SEG) hidden state when read by the same probe, that it gains nothing from the image where the hidden state does, and that it is concentrated in the late layers. Building on these findings, we introduce AttnSeg, which replaces the (SEG) embedding with this attention as the decoder's prompt. A learned, signed combination of the attention heads forms a dense spatial prompt, and an attention-guidance objective supervises it directly. A two-stage recipe first trains the VLM to localize without the decoder, then trains the decoder to refine, within the budget of standard fine-tuning. Applied to Sa2VA, AttnSeg improves ReasonSeg by 6.4 gIoU and GS-Eval by 10.2 at 4B (4.2 and 9.6 over the same model fine-tuned on the same data), and by 8.8 and 11.3 at 14B. The gains reach 13 to 18 gIoU on object parts, multiple objects, and single objects, where the referent must be singled out, and are negligible on stuff regions, which the decoder already recovers unaided. Code and models are available at https://eslambakr.github.io/attseg_project_page/

---

### On the Monitorability of Natural Language Autoencoders: Identification vs Discrimination

Authors: Kevin Yandoka Denamganaï
Published: 2026-10-13
Categories: cs.LG, cs.CL
arXiv: https://arxiv.org/abs/2610.18727v1

Natural Language Autoencoders (NLAs) are proposed as oversight tools for language models: an Activation Verbaliser (AV) writes an English explanation of an activation vector, and an Activation Reconstructor (AR) maps the explanation back to a vector. Their value for oversight depends on a human monitor being able to tell a true account of an activation from a false one, given the Av's explanation. We formalise NLA training and NLA oversight deployment as two different referential games from Emergent Communication: training is a generative referential game, while oversight is a discriminative, multi-modal, partially-observable referential game played over a many-to-many relation between the monitored language model's activations and the accounts that are true of them. The gap between the two games predicts that AV's explanations will identify which activation they describe without discriminating its true account from closely matched false ones, since training never asks explanations to rule anything out. Using the NLA's own AR as a maximally-sympathetic surrogate monitor model, we experimentally confirm this theoretical insight on one released checkpoint: true accounts beat accounts of other activations in all 320 comparisons, but beat fluent rivals differing in a single claim only 60.2% of the time (std.err. 6.7%, chance at 50%, $n = 35$). A meaning-preserving rewrite by an uninvolved model costs only 2.1% of the distance to another activation's account, so the failure is not due to the NLA pair having developed private conventions in the AV's explanations. Our rival accounts test only transcript-level truths, which the monitored language model's conversation settles, so these results bound the easiest case and leave claims about the monitored model's own processing untested. We state conjectures that would make both results consequences of the framework.

---

### Code-Switching Detection via Language-Selective Dimensions in Speech Foundation Models

Authors: Jie Chi, Zakaria Aldeneh, Masha Fedzechkina, Maureen de Seyssel, Natalie Schluter
Published: 2026-10-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.18726v2

Code-switching (CS), the alternation between languages within an utterance, is common in multilingual speech but underrepresented in pretraining corpora. Increasing CS coverage requires finding relevant utterances in large speech collections so that they can be included in training data. One way to find such utterances is to transcribe the speech and detect language alternation in the text, but this depends on obtaining accurate multilingual transcriptions. We propose a lightweight method that detects CS directly from speech. It uses frame-level language information in pretrained speech encoders to determine whether both languages are present within an utterance. Specifically, we identify language-selective hidden-state dimensions using monolingual recordings and aggregate their frame-level activations into a CS prediction. With a small CS set used to tune a decision rule, we show that the Whisper-small encoder achieves a mean F1 of $0.798$ across 14 language pairs in CS-FLEURS, compared with $0.776$ for a CNN and $0.751$ for a HuBERT linear baseline trained for monolingual language identification. Our findings suggest that pretrained speech encoders already encode the information needed to identify CS utterances, and they provide a lightweight tool for building more representative multilingual corpora.

---

### Code-Switching Detection via Language-Selective Dimensions in Speech Foundation Models

Authors: Jie Chi, Zakaria Aldeneh, Masha Fedzechkina, Maureen de Seyssel, Natalie Schluter
Published: 2026-10-13
Categories: cs.CL
arXiv: https://arxiv.org/abs/2610.18726v1

Code-switching (CS), the alternation between languages within an utterance, is common in multilingual speech but underrepresented in pretraining corpora. Increasing CS coverage requires finding relevant utterances in large speech collections so that they can be included in training data. One way to find such utterances is to transcribe the speech and detect language alternation in the text, but this depends on obtaining accurate multilingual transcriptions. We propose a lightweight method that detects CS directly from speech. It uses frame-level language information in pretrained speech encoders to determine whether both languages are present within an utterance. Specifically, we identify language-selective hidden-state dimensions using monolingual recordings and aggregate their frame-level activations into a CS prediction. With a small CS set used to tune a decision rule, we show that the Whisper-small encoder achieves a mean F1 of $0.798$ across 14 language pairs in CS-FLEURS, compared with $0.776$ for a CNN and $0.751$ for a HuBERT linear baseline trained for monolingual language identification. Our findings suggest that pretrained speech encoders already encode the information needed to identify CS utterances, and they provide a lightweight tool for building more representative multilingual corpora.

---

### Learning Source Structure: Bayesian Prediction with Transformers

Authors: Meir Feder, Ruediger Urbanke
Published: 2026-10-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.18725v1

A transformer trained on sequences whose source parameters are drawn afresh each time can learn only what the sequences share: the source family and its prior. The Bayesian predictor that knows both, which we call informed Bayes, is the unique minimizer of the training loss and is computable from the family alone; we use it as a quantitative, falsifiable reference. Where earlier work compared transformers with Bayes on sources whose contexts can be learned independently, we build Markov sources with up to 64 symbols in which observations in one context inform prediction in another: product sources with known or hidden encodings, transition rows sharing an unknown base law and depth, and a mixture of three families. Our main result is that informed Bayes predicts, from the family and prior alone, how a trained transformer behaves in four tests that go beyond the learning curve: the loss that remains under a source outside the training family (predicted 2.237 bits, measured 2.309), the doubling of the extra loss when two independent components change rather than one (0.345 predicted, 0.350 measured), the benefit of other contexts when a token appears for the first time, and the depth at which copying an earlier continuation starts to pay. On the learning curves themselves, small transformers match informed Bayes when the shared structure is visible in the tokens, and hidden structure needs more layers and training, with gaps below 0.01 bits per token for the largest models. We explain why the imitation is possible and where it must fail: informed Bayes decomposes into sufficient statistics that attention selects and accumulates and an estimator, fixed by the prior, that the feed-forward network learns; training pays the complexity of informed Bayes under the architecture's prior, and finite-memory universal prediction bounds how closely a bounded network can reach it.

---

### AIRRI: Towards a Human-Readable, Machine-Resistant Reading Interface

Authors: Emory Salaun, Syed Ayon, Kaleel Mahmood, Shaun Wallace
Published: 2026-10-13
Categories: cs.CV, cs.HC, cs.LG
arXiv: https://arxiv.org/abs/2610.18724v1

The rapid advancement of image-to-text extraction systems and large language models (LLMs) poses new challenges for educational assessment by making assignment materials easy to transcribe and solve with minimal human effort. To address this problem, we present AIRRI (Artificial Intelligence Resilient Reading Interface), a framework for generating machine-resistant educational content, a dataset of higher-education assignments for evaluation, and an initial evaluation of the human readability of AIRRI machine-resistant text designs. AIRRI employs two categories of machine-resistance techniques. First, AIRRI applies image transformations including character spacing, word spacing, line spacing, character opacity variation, stripe overlays, and image noise. Second, AIRRI applies decision-based adversarial attacks: RayS, SurFree, and ADBA. We evaluate AIRRI against three image-to-text systems: Tesseract, EasyOCR, and the transformer-based GOT-OCR2 model. Our results show that AIRRI substantially degrades automated text extraction performance. AIRRI increased character error rates for EasyOCR (6.9% to 55.9%), Tesseract (2.3% to 100.0%), and GOT-OCR2 (30.7% to 48.1%). Word error rates increased for EasyOCR (24.9% to 96.7%) and Tesseract (10.7% to 100.0%). Among the evaluated decision-based attacks, ADBA consistently produced the largest degradation in image-to-text performance, while RayS and SurFree produced more moderate reductions. EasyOCR and Tesseract were generally more susceptible to both image-based and adversarial perturbations than the transformer-based GOT-OCR2 model. Our findings suggest that AIRRI can significantly reduce the reliability of image-to-text-to-LLM workflows, increasing the effort required for AI-assisted assignment completion.

---

### On Collusion-Free Teaching in Planar Graphs

Authors: Khazaliya Liana, Kuske Clemens
Published: 2026-10-13
Categories: cs.DM, cs.CC, cs.LG
arXiv: https://arxiv.org/abs/2610.18714v1

Non-clashing teaching minimizes teaching complexity among models satisfying the Goldman--Mathias collusion-avoidance criterion. While the Vapnik--Chervonenkis dimension (VCdim) measures the complexity of learning from samples, non-clashing teaching dimension (NCTD) measures the number of examples required when a teacher can select them. Whether NCTD is always bounded above by VCdim remains a fundamental open question in learning theory. In this paper, we investigate this question for the concept classes of closed neighborhoods in planar graphs. We prove a tight upper bound of two on their NCTD, thereby establishing the conjectured inequality for this family. For the positive variant NCTD+, in which teaching sets contain only positive examples, we obtain a tight upper bound of four. Finally, we show that deciding whether either NCTD+ or NCTD is at most one is NP-complete, even for concept classes of closed neighborhoods in planar graphs. In particular, despite the upper bound of two on NCTD for such classes, distinguishing between NCTD values of one and two remains NP-hard.

---

### Steertuning Policies from a Few Human Videos

Authors: Billy Yan, Isaac Tu, Kelly Lee, Blake Chang, Kenny Ma, Kanad Patel, Dhawal Kabra, Neer Patel, Kevin Wu, Lerrel Pinto, Irmak Guzey
Published: 2026-10-13
Categories: cs.RO, cs.AI
arXiv: https://arxiv.org/abs/2610.18711v1

Current recipes for robot learning require massive amounts of robot demonstrations to obtain robust and adaptive policies. In this work, we introduce Steertuning, a system that enables robots to learn new tasks in new environments from 5 human demonstrations in under 5 Steertuning consists of two parts. (1) We pretrain a 400M-parameter diffusion policy on our egocentric human dataset, which consists of 100 hours of densely annotated random interactions. This pretraining makes Steertuning robust to background differences and enables spatial and object generalization when deployed on new tasks. (2) Given 5 demonstrations of a new task, we finetune our model for only 2 minutes (2000 training updates) and steer the policy's actions during reverse sampling using the demonstration whose observation is closest to that of the deployment environment. Together with a carefully designed data processing pipeline for embodiment transfer, this enables Steertuning to achieve a 71% success rate across 12 tasks in 12 new environments, ranging from lab settings to kitchens in new homes. The entire process, from data collection for a new task to robot policy deployment, takes less than 5 minutes. Our full dataset, model weights, codebase, and robot videos will be made available at our website: https://steertuning.github.io.

---

### Unlocking Token Rewards via Training-Free Reward Attribution

Authors: Sitong Wu, Haoru Tan, Bin Xia, Xichen Zhang, Jingyao Li, Shaofeng Zhang, Xiaojuan Qi, Bei Yu, Jiaya Jia
Published: 2026-10-13
Categories: cs.AI
arXiv: https://arxiv.org/abs/2610.18704v1

In this paper, we propose an extremely efficient, training-free method to extract token-level reward signals directly from an existing deep reward model. Our core idea is to attribute the overall process reward to individual tokens by estimating each token's influence. This influence is defined as the change in the final macroscopic reward (e.g., the process reward) when a token is replaced with a semantically null token. Naively calculating this influence is computationally infeasible, requiring N forward passes through the PRM for an N-token sequence. We overcome this bottleneck by proposing a highly efficient gradient-based estimator. Specifically, we use a first-order Taylor approximation, which simplifies the influence calculation to the inner product of the difference between the token embedding and the null token embedding, and the gradient of the reward with respect to the token embedding. This requires only a single forward and backward pass. The resulting token-level rewards enable standard RL algorithms to perform precise credit assignment without requiring additional reward model training. Experiments on challenging reasoning benchmarks demonstrate that our method substantially improves policy optimization efficiency and enhances the generalization of LLM reasoning capabilities. Our P2T outperforms the outcome reward by +4.9% on MathVista for Qwen2.5-VL-7B-Instruct, and +11.5% on AIME24 for Qwen2.5-Math-7B. The code will be released via https://github.com/stonewst/P2T.

---

### UniW(A)M: A Unified Mixture-of-Experts Framework for World and World-Action Models

Authors: Jingbo Xu, Jianjie Fang, Zhuohang Li, Zihao Song, Zhaolu Wang, Xvyuan Liu, Xin Zhang, Haisheng Su, Yinzhou Tang, Chen Gao, Wei Wu, Yong Li
Published: 2026-10-13
Categories: cs.RO, cs.AI, cs.CV, cs.LG
arXiv: https://arxiv.org/abs/2610.18697v1

World models (WMs) and world-action models (WAMs) focus on future world prediction and executable action generation, respectively. Although these models learn shared physical dynamics and interaction semantics from embodied trajectories, they are typically trained in isolation, which restricts knowledge reuse between prediction and control. The central challenge is therefore to develop a unified learning framework that shares these dynamics while preserving the distinct modeling capabilities required by each paradigm. We propose UniW(A)M, a mixture-of-experts world-(action) model built upon diffusion Transformers that unifies action-conditioned world modeling and diverse WAM variants. Shared experts capture common dynamics, while atomic capability experts are reused across interfaces so that each interface keeps its own conditional pathway. At deployment, a plug-in adaptive interface router (PAIR) learns from a small subset of target-task chunk scenarios to select a suitable WAM interface for each execution chunk, balancing task success and execution efficiency while keeping the backbone frozen. We evaluate UniW(A)M on LIBERO, RoboTwin 2.0, and three real-world manipulation tasks, where it achieves 88.3% average real-world success versus 76.7% for the strongest baseline. On WorldArena, it leads action-conditioned world modeling on all three tasks. AC-WM co-training contributes to action generation on every action interface and PAIR raises OOD success from 63.3% to 76.7%.

---

### CryoTracer: A Synthetic-to-Real System for Filament Instance Tracing in Cryo-EM

Authors: Radosław Kuczbański, Wojciech Kozłowski, Mateusz Grzesiuk, Konrad Karanowski, Lukas Frey, Andrzej Rzepiela, Maciej Zięba
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18696v1

Particle picking in cryo-electron microscopy is commonly formulated as an object-detection problem in which individual particles are represented by bounding boxes, a natural representation for globular particles but a poor match for filaments that can extend across an entire micrograph. Existing methods reconstruct filament trajectories by linking local predictions in a separate post-processing stage, decoupling detection from instance formation and creating potential ambiguities at gaps and crossings. Directly preserving filament identity is particularly useful for polymorphic specimens, where segments originating from the same filament should remain associated. However, learning such instance-level representations is challenging because expert centerline annotations are scarce and costly to obtain. We introduce CryoTracer, a synthetic-to-real framework that addresses this limitation through procedural micrograph generation, transformer-based polyline prediction, and specimen-specific fine-tuning. Procedural pretraining exposes the model to diverse filament geometries, intersections, and imaging conditions before adaptation using a limited number of fully annotated experimental micrographs. CryoTracer represents filaments as ordered centerline polylines and optimizes their geometry using an orientation-invariant discrete Fréchet objective, enabling consistent learning regardless of tracing direction. Evaluation on established and newly expert-annotated datasets shows that CryoTracer improves instance-level filament recovery over existing particle-detection approaches. These results establish end-to-end polyline prediction as a practical alternative to detection-and-linking pipelines for recovering complete filament instances with limited micrograph annotation budgets. Code and data are available at https://github.com/cryotracer/cryotracer

---

### A Refresh Interval Does Not Fully Specify a Refresh Schedule: Learning-Rate-Dependent Placement in Kronecker-Preconditioned Language Model Training

Authors: Xinyu Wang, Zhiwei Zhai, Yu Gu, Zijun Yu, Sicheng Lyu, Jingrui Tian, Lei Ding, Ziyang Song, Xiao-Wen Chang
Published: 2026-10-13
Categories: cs.LG
arXiv: https://arxiv.org/abs/2610.18695v1

Kronecker-preconditioned optimizers recompute a cached curvature eigendecomposition only periodically, and implementations and adaptive methods control this through the refresh frequency. We show that frequency leaves part of a schedule free: at a fixed number of refreshes, when they occur changes held-out loss at the same compute. Training GPT-style models with the Mousse optimizer across model sizes, we move the middle refreshes of a four-refresh schedule while fixing the first and last. On FineWeb the two placements exchanged rank as the learning rate rose (later placement was better at low rates, earlier placement at high ones), so a single-learning-rate comparison can report either ranking. The reversal held under controls including the published optimizer internals, and placement still mattered at the published refresh density. The learning rate at which the preference flips fell with model size, consistent with a held-out smaller model, and shifted with training length. A spectral-width staleness score predicted the same ranking at every learning rate and missed the reversal. Refresh schedules should be reported in full and compared across the learning rates at which they will be used.

---

### GeoReader: Learning Latent Geometry for Spatial Reasoning in Vision-Language Models

Authors: Rong Li, Shijie Li, Jiangyong Huang, Lingdong Kong, Jiaming Zhou, Dongyue Lu, Zeying Gong, Baoxiong Jia, Junwei Liang, Siyuan Huang
Published: 2026-10-13
Categories: cs.CV, cs.AI, cs.LG, cs.MM
arXiv: https://arxiv.org/abs/2610.18694v1

Vision-language models (VLMs) struggle with metric and multi-view relationships that remain implicit in RGB observations. Geometry foundation models recover this structure from RGB, but existing methods either fuse the estimated features with visual tokens and learn to use them from question answering alone, or align VLM latents with geometry features. Capturing geometry, however, does not ensure its use: zeroing the aligned latents of a released feature-alignment model leaves its accuracy unchanged. We introduce GeoReader, a framework that learns geometric representations for the VLM that reads them. First, we train lightweight Geometric Readers to produce continuous input tokens, supervising them with geometry descriptions decoded by a frozen VLM, so that the Readers express geometry in a form the target model can already read. We then freeze the Readers to preserve this interface while fine-tuning the VLM on spatial QA, with geometry injection conditioned on the question. Geometry descriptions serve only as training targets: at inference, the VLM answers directly from the continuous geometric tokens without generating a geometry chain of thought. With geometry predicted from RGB images, our approach reaches 94.3 on MindCube-Tiny with Qwen3-VL-4B and, with a multi-source QA mixture, an average of 72.5 across six additional spatial benchmarks with Qwen3.5-9B. These results suggest that a useful geometric representation is one the receiving model can use, not merely one from which geometry can be reconstructed.

---

### Orientation Estimation of Cryo-EM Images for Molecules with Arbitrary Symmetry by Semidefinite Programming

Authors: Yunpeng Shi, Amit Singer, Jinxin Wang, Ruiyi Yang
Published: 2026-10-13
Categories: cs.CV, math.NA, math.OC
arXiv: https://arxiv.org/abs/2610.18691v1

This paper extends the non-unique games framework introduced in Bandeira et al. (2020) to address symmetric structures. Precisely, we consider the problem \( \begin{aligned} \underset{g_1,\ldots,g_n\in \mathcal{G}}{\operatorname{min}}\,\, \frac{1}{|S|}\sum_{i,j=1}^n \sum_{s\in S} f_{ij}(g_i s g_j^{-1}), \end{aligned} \) where $\mathcal{G}$ is a compact group, $S\subset \mathcal{G}$ is a finite subgroup of symmetry elements, and $f_{ij}:\mathcal{G}\rightarrow \mathbb{R}$ is a loss function. A notable example is common lines based orientation estimation of cryo-EM images for molecules with point group symmetry $S$, where $\mathcal{G}$ is the rotation group $\operatorname{SO}(3)$. With a semidefinite programming relaxation, we turn the problem into a convex program that is solved with the alternating direction method of multipliers. Unlike existing approaches which are tailored to specific symmetry groups, ours provides a unified computational framework for addressing all point group symmetries. We demonstrate the applicability of the proposed algorithm on the cryo-EM orientation estimation problem on both simulated and experimental datasets. In particular, our algorithm is effective in generating high-quality ab initio models from only a few class averages computed from experimental images.

---

### Just Respect the Semantics: Image-Token Compression via Pre-LLM Semantic Variable-Resolution Tokenization and In-LLM Relevance Pruning

Authors: Tianhao Fu, Austin Wang, Charles Chen, Roby Aldave-Garza, Victor Xie, Xiyue Jiao, Yucheng Chen, Martin Tin
Published: 2026-10-13
Categories: cs.CV, cs.AI, cs.LG
arXiv: https://arxiv.org/abs/2610.18690v1

Image-token compression must reduce computation without hurting the semantics. We introduce Varest, guided by semantic fungibility: tokens within a coherent semantic region can be summarized coarsely, while object boundaries and literal text require finer distinctions. Object and text localizers guide variable-resolution merging, preserving spatial coverage and finer detail at boundaries and text. Learned (REL) registers then prune visual tokens after early decoder processing. Training across budgets enables one checkpoint per backbone. Across three backbones and 13 benchmarks, Varest outperforms 15 baselines in mean performance retention at every evaluated budget, retaining 86.10-92.33% of full-token performance with only 5.56% of the original layer-averaged image-token count.

---

### RSIgen: Reciprocal Self-Improvement for Agentic Visual Generation

Authors: Yidi Liu, Haozhe Wang, Jie Xiao, Kai Zhu, Xueyang Fu, Zheng-Jun Zha
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18688v2

Agentic visual generation produces experience that can improve the native understanding (agent) and generation roles of a unified multimodal backbone, but task success alone does not determine which role should learn. Generator updates can also make the agent's action supervision stale. We introduce RSIgen, a task-centric reciprocal self-improvement framework linking learning allocation, supervision refresh, and recursive curriculum construction. Dynamic Responsibility Routing (DRR) uses matched interventions on evidence, inputs, and supported actions to select learning recipients and construct compatible training targets. After generator learning, RSIgen re-evaluates the relative utility of tested actions with branch inputs held fixed to refresh agent supervision; the retained roles then construct the next curriculum. Across BAGEL, Qwen-Image, and Wan-Image-7B, RSIgen improves AgentGen-Bench Full Overall9 by 3.2--3.3 points over a three-cycle SearchGen adaptation with action costs. Wan controls support allocation gains at matched role-level training exposure, refresh gains at a matched adaptation budget, and benefits from updated curriculum sources while recipients continue learning. In blind evaluation on 200 real-user requests, pooled tie-adjusted preference scores against the same baseline reach 70.0% for task completion and 64.9% for visual quality.

---

### RSIgen: Reciprocal Self-Improvement for Agentic Visual Generation

Authors: Yidi Liu, Haozhe Wang, Jie Xiao, Kai Zhu, Xueyang Fu, Zheng-Jun Zha
Published: 2026-10-13
Categories: cs.CV
arXiv: https://arxiv.org/abs/2610.18688v1

Agentic visual generation produces experience that can improve the native understanding (agent) and generation roles of a unified multimodal backbone, but task success alone does not determine which role should learn. Generator updates can also make the agent's action supervision stale. We introduce RSIgen, a task-centric reciprocal self-improvement framework linking learning allocation, supervision refresh, and recursive curriculum construction. Dynamic Responsibility Routing (DRR) uses matched interventions on evidence, inputs, and supported actions to select learning recipients and construct compatible training targets. After generator learning, RSIgen re-evaluates the relative utility of tested actions with branch inputs held fixed to refresh agent supervision; the retained roles then construct the next curriculum. Across BAGEL, Qwen-Image, and Wan-Image-7B, RSIgen improves AgentGen-Bench Full Overall9 by 3.2--3.3 points over a three-cycle SearchGen adaptation with action costs. Wan controls support allocation gains at matched role-level training exposure, refresh gains at a matched adaptation budget, and benefits from updated curriculum sources while recipients continue learning. In blind evaluation on 200 real-user requests, pooled tie-adjusted preference scores against the same baseline reach 70.0% for task completion and 64.9% for visual quality.

---

### RadarGSOcc: Doppler-Aware 3D Semantic Occupancy Prediction Using 4D Radar Tensor with Geometric-Semantic Feature Extraction

Authors: Christofel Rio Goenawan, Dong-Hee Paek, Seung-Hyun Kong
Published: 2026-10-13
Categories: cs.CV, cs.AI, cs.RO
arXiv: https://arxiv.org/abs/2610.18686v1

3D semantic occupancy prediction is crucial for autonomous driving, but camera-based methods degrade in low-light conditions and LiDAR-based methods are sensitive to adverse weather. 4D radar is more robust, yet its sparsity and the high cost of radar tensor processing limit its use. We propose Radar Geometric-Semantic Occupancy Prediction (RadarGSOcc), an efficient framework that directly processes 4D radar tensors by separately extracting geometric and semantic features, including Doppler information, with a deformable attention-based geometric refinement module as its key component. We also introduce the K-Radar 3DSemOcc benchmark, the first large-scale benchmark for 3D semantic occupancy prediction using 4D radar tensors, built on the K-Radar dataset with refined LiDAR-based annotations. RadarGSOcc predicts 15 classes within a 51.2m range and achieves 38.2% IoU, 7.6% mIoU, and 25.5% weighted mIoU, outperforming RadarOcc by 17% relative (25.5 vs. 21.9) in weighted mIoU while using fewer parameters and lower computation. It substantially improves truck prediction while trading off car accuracy, outperforms camera-based approaches, and narrows the gap with LiDAR-based methods. To assess robustness without per-frame ground truth in adverse weather, we further propose a novel proxy metric, Environment-Based KL Divergence (EBKL), under which 4D radar proves a reliable and efficient modality for 3D semantic occupancy prediction. The dataset and code will be publicly available.

---

### Knowing the User Is Not Knowing What We Established: Grounding Awareness in Personal Agents

Authors: Jiajun Chen, Xinyue Chen, Yicheng Jiang, Hua Shen
Published: 2026-10-13
Categories: cs.AI, cs.HC, cs.LG
arXiv: https://arxiv.org/abs/2610.18684v1

Personal-agent memory has largely focused on learning what users know, prefer, and want. However, knowing the user does not necessarily mean knowing what the user and the agent have established together. An assistant's proposal should not become an authorized instruction merely because it is remembered. We introduce Grounding Awareness, the ability to distinguish accepted, unacknowledged, revised, and scoped arrangements when selecting actions. We develop a controlled benchmark to test whether agents can tell the difference between proposals users have accepted and those they have not, and whether they apply accepted arrangements only where appropriate. Across multiple agents and memory systems, we find that strong memory performance does not ensure Grounding Awareness. More surprisingly, providing more complete user information can increase execution of unacknowledged proposals, while persistent memory can cause agents to treat these proposals as if they were user instructions. We trace these failures through memory writing, retrieval, and action selection. Finally, we introduce Grounding Layer, a lightweight, training-free extension that explicitly tracks grounding status and checks consequential actions, substantially improving agent behavior without modifying models or memory systems.

---

### ProAS: A Progress-Aware Advantage Shaping Framework for Multi-Turn Agentic Reinforcement Learning

Authors: Bowen Ding, Mingda Wan, Yuhan Chen, Jiyao Yuan, Qi Zhu, Futing Wang, Heyuan Deng, Fei Mi, Lifeng Shang, Tao Lin
Published: 2026-10-13
Categories: cs.LG, cs.AI
arXiv: https://arxiv.org/abs/2610.18683v1

Reinforcement learning (RL) has emerged as the primary paradigm for optimizing Large Language Model (LLM) agents, yet efficient critic-free algorithms (e.g., GRPO) suffer from a fundamental uniform credit assignment dilemma that broadcasts trajectory-level advantages equally to all actions. Our analysis reveals that this uniformity inherently reinforces structurally invalid actions in preferred trajectories and ignores the downstream cascading influence of early pivotal deviations in unpreferred ones. To resolve this without introducing critic overhead, we propose Progress-Aware Advantage Shaping (ProAS), a plug-and-play framework that leverages a dynamically evolving Agentic Reference Graph (ARG) to achieve fine-grained credit assignment via Masking Unreasonable Incentives (MUI) and Asymmetric Penalty Redistribution (APR). Extensive experiments across Qwen2.5-1.5B-Instruct and Qwen2.5-7B-Instruct demonstrate that ProAS universally elevates mainstream group-based baselines to new state-of-the-art performance, achieving peak success rates of 91.9\% and 94.3\% on ALFWorld, alongside 66.7\% and 70.8\% on WebShop, respectively. Furthermore, it accelerates training convergence by up to $2.44\times$ while maintaining strong robustness against imperfect or distilled ARGs. Codes: https://github.com/LINs-lab/ProAS.

---