# Research Papers — 2026-09 · bundle page 212 of 266 > Full-text markdown of papers #10551–10600 of 13298 published in 2026-09, newest first. AI-readable mirrors of the full 13298-paper 2026-09 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-09/page/212 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/211.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/213.md) ### Agentic ML Exploration (A-MLE) for Ads Ranking Authors: Erwin Gao, Vinodh Kumar Sunkara, Jingyi Guan, Qinjin Jia, Hangjun Xu, Xiang Ji, Sherman Wong, Surya Teja Chavali, Pratik Vaishnavi, Aryan Pandhi, Xiaoyu Deng, Zhaodong Wang et al. Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08248v1 Modern industrial ads ranking stacks are increasingly bottlenecked not by model capacity or training compute, but by the throughput of human ML iteration - the cycles of research, implementation, training, debugging, evaluation, and launch required to surface a single statistically significant improvement. A typical ranking stack contains numerous differentiated models with heterogeneous data, architectures, and infrastructure constraints, and each cycle takes days to weeks of senior engineer attention per model. As a result, techniques that have proven effective on one model diffuse into others slowly and unevenly, leaving substantial recoverable signal unexplored. We present Agentic ML Exploration (A-MLE), an autonomous LLM-agent system that systematically explores ML techniques across a portfolio of ads ranking models. A-MLE decomposes ML iteration into five stages involving hypothesis generation, exploration strategy, experiment execution, result analysis and shared knowledge substrate which are orchestrated by a single agent that invokes domain-specific skills and agentic workflows against a sandboxed execution layer, with human-in-the-loop checkpoints at each stage boundary. We deploy A-MLE across a representative set of large-scale ads ranking models and evaluate it along a tiered capability framework (tool availability, autonomous workflow execution, and open-ended exploration). We further report a controlled cross-LLM study using a fixed agent loop, which surfaces qualitative differences in execution reliability and exploration aggressiveness across the Claude Sonnet, Gemini, and GPT families. We discuss failure modes and the design choices that govern reliability. Our findings suggest that agentic exploration is a practical force multiplier for ML engineers in industrial recommenders, especially for the long tail of models that rarely receive expert attention. --- ### zScore-N: A Neural Network for On-Chain Wallet Reputation Scoring Authors: Girish G N, Ashutosh Sahoo, Akshay SP, Gurukiran S, Dhanashekar Kandaswamy Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08247v1 Wallet reputation scores decide who receives an airdrop, who can borrow, and who enters an allowlist across decentralised finance. They almost always begin as hand-written formulas: compositions of clamped logarithmic, linear and square-root transforms over behavioural features, with every threshold and point award set by hand. Such a formula is readable and deterministic, but it is piecewise and non-differentiable, it cannot improve as data accumulates, and it cannot distinguish a feature that is genuinely zero from one its pipeline failed to capture. We present zScore-N, the neural network that replaced ours in production. The formula served as its teacher: calibrated against 5,208,952 wallets sampled across 2019-2024 and verified to reproduce production output to within 2.3e-13, it supplies unlimited labelled training data at zero label noise. The trained network reproduces it to 0.58 points RMSE on the 1000-point scale (R^2 = 0.99997), against 2.25 for gradient-boosted trees and 28.04 for linear regression on identical features and splits. Trained with missing-value masks against uncorrupted targets, it halves the error that incomplete data introduces: at 10% feature-level missingness the formula drifts 51.4 points from its own complete-data output with a systematic -12.5 point bias, while the network drifts 17.9. The network carries the score at production scale, across a population of millions of wallets spanning six orders of magnitude in size and activity. --- ### CUNO: Curriculum and Preference Optimization for Stable Graph Unlearning under Mass Deletion Authors: Chenhan Zhang, Ali Braytee, Madhushi Bandara, Xin Hao, Paul J. Kennedy, Massimo Piccardi, Raymond Owen Published: 2026-09-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.08244v1 Graph unlearning removes the influence of designated training data from a trained graph model without retraining from scratch. However, existing methods suffer a sharp drop in model utility under large deletion ratios (mass deletion), a phenomenon we refer to as catastrophic unlearning. We find that a key cause is the uniform treatment of all deleted samples, which is particularly damaging in graph learning: structural dependencies cause different nodes to play vastly different roles in the learned model, yet existing methods apply the same forgetting operation to the entire forget set. Based on this insight, we propose CUNO, a curriculum-based graph unlearning framework that removes the forget set progressively, ordering samples by their estimated unlearning difficulty across multiple stages. CUNO further employs a distribution-level negative preference optimization (NPO) objective at each curriculum stage that steers the model away from its original behavior on the current forget subset while preserving retained performance. Our theoretical analysis shows that the curriculum design is most beneficial when the forget set spans a wide range of unlearning difficulty, a condition naturally satisfied under mass deletion. Comprehensive experiments confirm that CUNO consistently mitigates catastrophic unlearning: at 20% deletion, it retains 74% of the original utility compared to 26-53% for existing methods, and maintains more than half the original utility even at 50% deletion. Our code is publicly available at https://anonymous.4open.science/r/cuno-D4FF. --- ### CS-CLIP: Compositional Scene Graph-guided CLIP for Robust Compositional Reasoning Authors: SeongJun Jeong, Minjoon Jung, Woo Suk Choi, Youwon Jang, Byoung-Tak Zhang Published: 2026-09-08 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.08242v1 Vision-language models (VLMs) demonstrate strong performance across compositional reasoning benchmarks, which require reasoning over semantic perturbations of objects, attributes, relations, and their interactions. However, our controlled analysis reveals that existing compositionality-aware VLMs exhibit element-specific biases, often underperforming vanilla CLIP on certain compositional elements. To address this, we propose Compositional Scene Graph-guided CLIP (CS-CLIP), which uses scene graphs to identify compositional elements and construct structured negatives via selective masking. We further retain negatives that are most contradictory to the original caption, forcing the model to rely on compositional structure rather than surface cues. CS-CLIP achieves state-of-the-art compositional reasoning with robust performance across compositional elements. It also preserves general vision-language capabilities such as cross-modal retrieval and downstream visual reasoning, while requiring fewer training samples than prior methods. --- ### Style Over Substance: Content-Invariant Wrappers Flip LLM Safety-Judge Verdicts Authors: Yongxi Zhou, Wenbo Ye, Yuanzhe Liu, Zihan Dong, Junwei Yao Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08236v1 Automatic safety judges -- systems such as Llama Guard or a GPT-4o grading prompt that decide whether a model's reply is harmful -- produce the numbers behind almost every reported jailbreak success rate, defense evaluation, and safety leaderboard. We ask whether these judges grade what a reply contains or how it sounds. We keep a reply's content fixed and add content-invariant style wrappers: fixed strings placed before or after the reply that change only its tone (an educational disclaimer, a fake safety "reasoning" block, a token refusal followed by the unchanged harmful body), or, on harmless refusals, framing that merely sounds dangerous. The body is preserved byte-for-byte, so a faithful judge must return the same verdict, and any flip is an error of the judge, not a change in safety. Over 600 JailbreakBench replies x up to 7 forms x 8 judges, we measure flip rates with paired significance tests and measured noise floors. Findings are precise rather than universal: most judges barely move, but specific judges harbor cheaply exploitable blind spots. A token-refusal wrapper flips 19.9% of GPT-4o-mini's correct "unsafe" verdicts (noise floor 0.5%; 18.2% under majority-of-three re-scoring) yet moves Claude only 0.4%. The deployed Llama Guard 4 is deterministically gamed: an "educational course" framing flips 12.3% of its harmful verdicts to safe. A second deployed guard (gpt-oss-safeguard-20b) is immune, and rewriting only the grading prompt (StrongREJECT-style) cuts the attack tenfold on the identical model -- the vulnerability lives in the judge, not the content. A two-annotator human validation confirms 100% content invariance and 90% of flips as judge errors (kappa 0.95-1.0), and a bootstrap shows the underlying model ranking is already unstable to sampling alone. We release the dataset, wrappers, code, and per-verdict labels. --- ### Distribution-free inference on the number of changepoints Authors: Rohan Hore, Aaditya Ramdas Published: 2026-09-08 Categories: stat.ML, cs.LG, stat.ME arXiv: https://arxiv.org/abs/2609.08234v1 Suppose we are given an ordered sequence of independent data whose distribution changes $K$ times at unknown locations, for some unknown $K \geq 0$. In this paper, we study the problem of performing distribution-free inference on $K$. First, we show an impossibility result: any distribution-free upper confidence bound on $K$ must be trivial and uninformative. Then, using conformal $p$-values, and under only the assumption that the data segments induced by the changepoints are exchangeable (within themselves) and mutually independent, we construct a finite-sample valid lower confidence bound on $K$, which we call the Conformal LOwer bound on Changepoint Count (CLOCC). We show that CLOCC is the only feasible way to provide a lower bound on $K$ under the stated assumptions, a property we refer to as its universality. We provide practical guidelines for choosing score functions that yield efficient and tight lower bounds. We evaluate CLOCC in several synthetic and real-data experiments, where it provides informative lower bounds on $K$, demonstrating its practical applicability. --- ### Routing Dense Layouts with History-Aware Offline Reinforcement Learning using LSTM Authors: Afsara Khan, Austin Rovinski Published: 2026-09-08 Categories: cs.AR, cs.LG arXiv: https://arxiv.org/abs/2609.08232v1 Detailed routing remains a dominant runtime bottleneck in physical design due to increasing complexity of design rules. Modern routers can struggle to resolve persistent violations under dense operating conditions. While recent work leverages reinforcement learning (RL) to dynamically select costs for each routing iteration, we find that this technique struggles with high-density designs where routing solutions are significantly harder. To address this, we present a history-aware offline RL policy which predicts iterative cost weights in these dense regimes to improve convergence across placement densities by utilizing readily available features from the router. Our policy uses conservative Q-learning similarly to prior work; however, our key insight is that addition of a lightweight LSTM architecture and additional features can retain sequence context and improve routing convergence across multiple densities and route guide qualities. Our policy can be integrated into any cost-based router with minimal pipeline changes, as it does not interfere with the core search algorithm. We evaluate our policy on held-out density and adjustment settings, including difficult operating points induced by dense placement and low guide quality. Our policy reduces design rule violations (DRVs) by an average of 92% over the top public baseline while simultaneously reducing runtime by 10%. --- ### ActionSplice: In-Flight Action Editing for Interactive World Models Authors: Pardis Taghavi, Tingyu Guo, Jonas Lossner, Gaurav Pandey, Reza Langari Published: 2026-09-08 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2609.08230v2 Chunk-autoregressive video world models typically generate each chunk under one action. When an action changes during sampling, waiting until the next chunk delays the response, while directly switching the conditioning leaves the intermediate solver state shaped by the previous action. Restarting sampling under the revised action avoids this mismatch but repeats completed computation. We introduce \emph{ActionSplice}, which edits the interrupted state through Counterfactual State Transport (CST). A lightweight corrector moves the interrupted solver state toward the matched counterfactual solver state induced by the revised action at the same solver step. The world model and sampler remain frozen, and sampling resumes without replaying completed evaluations. $\mathrm{CST}_{R}$ retargets the entire active chunk, while $\mathrm{CST}_{T}$ preserves the temporal prefix at the intervention step and corrects only the suffix. On minWM--Wan Action2V and HY-WM1.5, we measure fidelity against matched rollback references. Relative to condition swapping, $\mathrm{CST}_{R}$ reduces LPIPS by 61.5\% and 75.9\%, and $\mathrm{CST}_{T}$ reduces suffix LPIPS by 56.1\% and 77.5\%, respectively. --- ### SE-GoS: Self-Evolving Graph-of-Skills for Skill Library at Scale Authors: Dawei Fu, Cheng Jiang, Sitian Qian, Huainan Wang, Zhongkai Hao Published: 2026-09-08 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.08228v1 Modern LLM agents increasingly rely on reusable skills, yet as skill libraries scale to thousands of entries, effective retrieval becomes a bottleneck. Graph-of-Skills (GoS) addresses this challenge by exploiting dependency-aware graph structure for scalable skill retrieval, while SkillDAG further demonstrates that skill graphs can accumulate execution-backed structure online. However, these approaches leave open whether historical execution traces can be systematically distilled into a better retrieval graph that generalizes to unseen tasks. We present Self-Evolving Graph-of-Skills (SE-GoS), a training-free framework that evolves an existing GoS graph from execution traces while preserving the original retrieval pipeline. SE-GoS performs three complementary updates: topology evolution that discovers and prunes skill relationships from execution evidence, edge-weight evolution that reinforces retrieval-relevant relationships based on historical effectiveness, and description evolution that optimizes retrieval-facing skill descriptions using execution feedback. Across three LLMs on SkillsBench, SE-GoS consistently improves task reward while reducing input tokens relative to full skill loading, with gains varying across model families. In a representative setting, one evolution round improves reward from 52.4\% to 59.4\% while reducing input tokens by approximately one-third relative to full skill loading, and the resulting graph transfers to a disjoint held-out split with a 5.4-point improvement over the static GoS baseline. These results show that skill graphs can be improved from execution experience without model training, changes to the retrieval algorithm, or modifications to skill content, turning a static retrieval graph into an evolving retrieval infrastructure. --- ### TTGBench: Benchmarking Topological Evolution and Semantic Drift in Text-attributed Temporal Graphs Authors: Longfei Ma, Zemin Liu, Fei Wu Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08226v1 Temporal graph learning models the evolution of dynamic systems, where both structural interactions and semantic states change over time. However, existing benchmarks primarily emphasize structural evolution via temporal link prediction (TLP), while support for semantic evolution remains limited. Although temporal node classification (TNC) is sometimes included, it is typically restricted to simplistic binary settings that fail to capture realistic semantic drift. Moreover, commonly used datasets exhibit high link repetition, leading to inflated performance estimates and obscuring true model capability. To address these limitations, we introduce \textbf{TTGBench}, a new benchmark that jointly evaluates structural and semantic evolution. TTGBench comprises six real-world, text-rich datasets characterized by \emph{Dual Volatility}, enabling rigorous and fair evaluation of existing models. Notably, it is the first benchmark to support both multi-class and multi-label TNC, filling a critical gap in evaluating temporal semantic drift. We conduct a comprehensive evaluation of 17 state-of-the-art methods across Temporal Graph Neural Networks (TGNNs) and Large Language Model (LLM)-based paradigms. The results reveal a clear \emph{capability divide} between the two paradigms: TGNN-based methods excel at structural prediction but fail at semantic tracking, whereas LLM-based predictors show the opposite trend. Through in-depth analysis, we uncover their fundamental limitations and provide insights for developing more comprehensive temporal graph models. --- ### 3DWay: Generalizing Robot Manipulation via 3D Consistent Waypoints Authors: Ziqin Huang, Yingyue Li, Chenyangguang Zhang, Ruida Zhang, Yuxin Chen, Gu Wang, Xingyu Liu, Masayoshi Tomizuka, Xiangyang Ji Published: 2026-09-08 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.08224v1 Intermediate representations are key to bridging the modality gap between generalizable manipulation policies and large-scale pretrained vision-language models (VLMs). Among these, trajectory-based representations compactly represent motion-relevant cues, yet most existing approaches predict trajectories in 2D image space, resulting in intrinsic 3D ambiguity. Moreover, using 2D trajectories with depth still leaves the free-space waypoints ambiguous, limiting reliable 3D reasoning. To address this, we propose predicting 3D consistent waypoints (3DWay) from multi-view images. By reformulating 3D waypoints prediction as generating multi-view consistent 2D waypoints followed by geometric triangulation, we enable explicit 3D motion specification while preserving the strong priors of pretrained VLMs. The predicted waypoints can guide existing VLA models for better generalization or be directly executed on simple tasks. Extensive experiments show that 3DWay substantially improves 3D spatial grounding and vision-language reasoning, demonstrating strong potential for generalizable robot manipulation. Codes will be released at https://github.com/ziqin-h/3DWay. --- ### SoftRerank: Hierarchical Soft Fusion with Candidate-Label Reranking for Long-Tailed Micro-Action Recognition Authors: Yichi Zhang, Zhichao Xia, Yanjun Chi, Lingsi Zhu, Yuefeng Zou, Jun Yu, Qingsong Liu, Jianqing Sun, Shengping Liu Published: 2026-09-08 Categories: cs.CV, cs.MM arXiv: https://arxiv.org/abs/2609.08221v1 Micro-actions are subtle, low-intensity non-verbal behaviors that provide cues to fine-grained human states, including emotions and intentions. Recognizing them remains difficult because they are brief, contain weak visual changes, and often exhibit similar motion patterns across categories. This paper addresses these challenges with a fine-grained micro-action recognition method that combines full fine-tuning of InternVideo2.5, hierarchical soft fusion, and a lightweight candidate-label reranker. For the long-tailed label distribution in MA-52, we use class-balanced sampling and inverse-frequency reweighting to reduce the effect of frequent classes during training. We fine-tune InternVideo2.5 end to end and attach coarse and group-conditional fine-grained classification heads to the shared video representation, improving the consistency between coarse and fine predictions. For ambiguous samples, the candidate-label reranker uses hard samples and video-label matching to focus on easily confused fine-grained actions. Experiments validate the proposed method, which achieves a 79.99% F1-mean on MA-52 and ranks first in the 3rd Micro-Action Analysis Grand Challenge at ACM Multimedia 2026. --- ### Speed Limit for Information Acquisition in Stochastic Learning Dynamics Authors: Shuta Kobayashi, Andreas Dechant Published: 2026-09-08 Categories: cond-mat.stat-mech, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.08219v1 Neural networks acquire internal representations through learning. In this work, we formulate stochastic gradient descent (SGD) as a Markovian stochastic process and derive a Fisher-information flow speed limit that bounds the rate at which trainable parameters can acquire information about latent variables in the data-generating process. The resulting inequality decomposes the information flow into drift and noise contributions, thereby quantifying the roles of deterministic learning forces and SGD-induced fluctuations from an information-theoretic perspective. We verify the bound in analytically tractable basis-function linear regression, where the information budget predicted by the bound reproduces the ordering and characteristic time scales with which different latent variables are encoded in the learned parameters. These results establish Fisher-information speed limits as a quantitative framework for diagnosing when and how different aspects of the data-generating mechanism are acquired during stochastic learning. --- ### Drive by Hindsight and Foresight: Tool-Grounded Synergistic Reasoning over Hierarchical Memory for Autonomous Driving Authors: Baojie Chen, Zijun Jia, Jing Zhong Published: 2026-09-08 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.08217v1 VLMs have shown promise for autonomous driving, yet still suffer from hallucination, weak spatio-temporal perception, and limited generalization. Recent methods improve reasoning and decision-making through CoT explanations, retrieval-augmented generation or the static injection of tool outputs. Although these mechanisms enrich the context, the model neither proactively perceives scene information nor accumulates experience after answering. To overcome these limitations, we present, to our knowledge, the first synergistic framework that tightly couples hierarchical memory with proactive tool invocation in a closed reasoning loop. Our contributions are threefold. (i) Hierarchical Driving Memory: a scene-level short-term memory maintains the dynamic scene state, and an evolving long-term memory retrieves reusable experience and tool strategies. (ii) Memory-Tool Synergistic Reasoning Framework: guided by the scene state and retrieved experience, the model adaptively invokes tools to refine its reasoning at inference time and consolidates reusable experience into a long-term memory pool offline. (iii) Data Generation and Two-stage Training Pipeline: verified memory-tool trajectories built by multi-step teacher rollout are used to train with SFT and GRPO. Our 7B model reaches an overall reasoning score of 80.03 and MCQ accuracy of 79.09% on DriveLMM-o1, surpassing the strongest baseline by 7.74 MCQ points and generalizes strongly across benchmarks. Notably, ablation and analysis studies validate the effectiveness of each component and further reveal the complementary roles of hierarchical memory. Short-term memory strengthens spatio-temporal understanding, improving STSBench accuracy by 24.2 points, while offline long-term memory consolidation yields an additional 3.57-point MCQ gain with all parameters frozen, demonstrating continual self-evolution through accumulated driving experience. --- ### Vision: Data-Centric Anchoring for Robust and Interpretable Agentic AI Authors: Arun Vignesh Malarkkan, Xinyuan Wang, Yanjie Fu Published: 2026-09-08 Categories: cs.AI, cs.ET, cs.LG arXiv: https://arxiv.org/abs/2609.08216v1 Agentic AI systems built on large language models fail in two persistent ways that scaling does not fix: they break under distribution shift, and they cannot explain the decisions they make. We argue these are co-symptoms of one structural deficiency in the data lifecycle that governs how agents are trained, evaluated, and deployed. Observational interaction logs record what an agent did, not what it would have done otherwise. They encode spurious correlations without controlled variation, so they lack the counterfactual structure needed to separate causal signal from coincidence or to validate an explanation. No model-centric method can recover invariances the data never contained. We present Data-Centric Anchoring: robustness and interpretability should be engineered into the data environment, not extracted from models after training. Our central contribution is the Data-Centric Agentic Loop, a four-stage framework of Curate, Augment, Constrain, and Attribute. The ordering is structural, not stylistic. Curation precedes augmentation because generative models amplify whatever bias they are trained on. Augmentation precedes constraint because invariance objectives are vacuous without variation across environments to be invariant to. Attribution closes the loop, converting observed failures into targeted data interventions for the next iteration. Each stage manufactures the preconditions of the next, which makes the loop self-correcting rather than merely sequential. We ground the framework in a failure-driven taxonomy that links four core failure modes to the data lifecycle: spurious feature reliance, distribution-shift fragility, uncertainty miscalibration, and explanation unfaithfulness. We close with the limits of this approach and the open problems that stand between it and practical deployment at scale. --- ### PhysFlow: Physics-Aware Optical Flow for Motion Controllable Video Generation Authors: Cong Wang, Hanxin Zhu, Yonglin Tian, Jiayi Luo, Ruiqi Song, Boyi Sun, Long Chen, Zhibo Chen Published: 2026-09-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.08215v1 Video generation models have recently attracted substantial attention for their ability to generate visually compelling videos, yet ensuring physically consistent and plausible dynamics still remains a fundamental challenge, driving a growing line of research on physical realism in video generation. To address this challenge, motivated by the fact that physical regularities are primarily encoded in motion patterns, we propose PhysFlow, a novel two-stage framework for improving the physical plausibility of generated videos by decomposing video generation into motion-aware optical flow generation followed by motion-conditioned appearance synthesis. Specifically, PhysFlow consists of a physics-aware optical-flow video generator called PA-Flow and a flow-guided video generator called FlowRender. During the first stage, PA-Flow employs a physics-aware attention module to model how motion attributes and material properties influence global motion and local deformation, respectively, and generates an optical flow video as an explicit representation of motion. In the second stage, FlowRender leverages the decoupled motion representation as guidance to synthesize realistic textures and appearances, ultimately producing the final physically plausible video. To further support model training with explicit physical supervision, we construct PhysVideo, a physics-based video dataset generated with a physics engine and 3D-GS rendering, containing 10K foreground objects and 50K realistic video sequences with annotations of motion and material properties. Extensive experiments demonstrate that our proposed PhysFlow generates videos with superior physical plausibility while maintaining high visual fidelity compared with existing methods. --- ### DRIFT: Removing Diffusion Watermarks by Deflecting the Generative Trajectory Authors: Rui Bao, Zheng Gao, Xiaoyu Li, Xiaoyan Feng, Yang Song, Jiaojiao Jiang Published: 2026-09-08 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2609.08213v1 Diffusion watermarking embeds verifiable signals into the generative process and commonly verifies them by recovering trajectory-dependent evidence, making the marks robust to conventional pixel-space distortions. Existing removal attacks either regenerate along deterministic trajectories, which often preserve the watermark-bearing latent structure, or optimize every image separately. We identify the reliance on a recoverable generative trajectory as a common attack surface among the schemes we study. Based on this observation, we propose DRIFT, a black-box attack that combines partial forward diffusion with stochastic reverse resampling. Forward re-noising limits source information available to a fixed-depth recovery pipeline, while stochastic reversal supplies alternative noise-driven paths whose removal benefit we isolate through matched sampler comparisons. Adaptive DRIFT searches a selected ladder for each image's first verifier-rejected rung and refines fidelity while retaining only updates rejected by the same verifier. At fixed depth, we derive information-theoretic and Wasserstein source-dependence bounds; under realized-ladder monotonicity, the first rejected rung is least distorted among rejected rungs on that ladder, and verifier-gated refinement preserves rejection. Across nine watermarks spanning three paradigms, DRIFT achieves 98-100% attack success and the best image quality among the compared attacks, without secret keys, verifier internals, or per-image gradient optimization. --- ### A Better Spur Should Start From Each Objective Authors: Shanwen Mao, Hao Zhang, Guangtao nie, Zhiheng Li, Huimu Wang, Sulong Xu, Gu Simiu Published: 2026-09-08 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2609.08211v1 Real-world Multi-Objective Reinforcement Learning (MORL) often suffers from sparse rewards, reward conflicts, and late-stage reward tug-of-war, causing traditional linear scalarization to experience severe metric oscillations. To address optimization conflicts among multiple objectives in real-world deployment scenarios, we propose Multi-Marginal Preference Optimization (MMPO), a fine-grained framework that intervenes at the data, gradient, and constraint levels rather than relying on coarse-grained global scalarization. Specifically, MMPO performs exposure debiasing to mitigate sparse and biased rewards, applies priority-aware orthogonal projection to decouple conflicting gradients, and introduces self-prompted gradient constraints to prevent dominant objectives from overwhelming weaker ones. Experiments on real-world e-commerce datasets show that MMPO improves training stability and consistently achieves better performance across conflicting metrics. Moreover, it generalizes robustly to broader tasks such as ToolRL and code generation, demonstrating its effectiveness as a practical paradigm for multi-objective alignment. --- ### SIM: Subspace Interaction-based Method for Token-Level Text Anomaly Detection Authors: Kehan Yan, Yue Tan, Qingfeng Chen, Shiyuan Li, Yu Zheng, Yixin Liu Published: 2026-09-08 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.08200v1 Token-level text anomaly detection, as an emerging trend of text anomaly detection, moves beyond coarse-grained document-level detection by localizing anomalous tokens within text. By providing fine-grained abnormality prediction, token-level text anomaly detection plays a critical role in various real-world applications, such as spam filtering and fake news detection. However, existing methods still rely on the global distance calculation for scoring, during which the local anomaly signals are severely diluted by numerous redundant normal feature dimensions. Moreover, pre-trained language models used in these methods inevitably smooth out surface anomalies, further limiting their effectiveness in token-level anomaly detection. To address these limitations, we propose a Subspace Interaction-based Method (SIM for short) for token-level text anomaly detection. To prevent local signal dilution, SIM adopts a subspace interaction-based anomaly detector, which decouples high-dimensional token embeddings into multiple low-dimensional ones, amplifying localized anomaly signals hidden within specific dimensions. To counteract the over-smoothing effect, we design a hard pseudo-anomaly generation module to construct pseudo-anomalous tokens, simulating the subtle anomalies obscured by semantic smoothing. Also, a probabilistic boundary loss is developed to standardize anomaly scores into statistical distances, effectively enforcing anomalous instances to deviate significantly from the normal distribution center. Extensive experiments on multiple benchmark datasets verify the effectiveness of SIM and demonstrate its remarkable efficiency, robustness, and interpretability. The source code is available at: https://github.com/yankehan/SIM-TAD. --- ### Qiushi Engine on AstaBench E2E-Bench-Hard Authors: Wenhao Li, Shuxing Yang, Fujia Chen, Jincheng Mi, Yuang Pan, Rui Zhao, Zichen Li, Junyao Wu, Shenzhan Hong, Yaqi Li, Yize Wang, Kaihao Zhu et al. Published: 2026-09-08 Categories: cs.AI, cs.SE arXiv: https://arxiv.org/abs/2609.08196v1 This report analyzes Qiushi Engine v0.8 across all 40 test tasks in AstaBench E2E-Bench-Hard, a benchmark that requires autonomous agents to carry a research question through experimental design, code implementation, actual execution, result analysis, and report delivery. Qiushi Engine is model-configurable; this evaluation selected DeepSeek deepseek-v4pro-preview as the model backend. The official AstaBench leaderboard records a score of 0.816 and an average benchmark cost of USD 15.209 per task, while the full-precision local recomputation is $81.59 \pm 1.87$. Four tasks satisfied every rubric item, yielding a full-task completion rate of 4/40 = 10% -- 7 percentage points above, and about 3.3 times, the approximately 3% best rate reported for AstaBench's official agents. Across 507 required rubric items, 416 were satisfied (82.1%). Official scoring archives and 40 Meta-Trace records show sustained production and verification of reports, code, and experimental artifacts; the principal gaps lie in repeated runs, external dependencies, specified metrics, and ablation studies. The report explains the benchmark, system workflow, aggregate results, representative cases, and limits of interpretation. --- ### Do Dynamic Routers Need Memory? HeRo: History-Aware Routing for Efficient LLM Inference Authors: Hongjin Lin, Wentao Wan, Keze Wang Published: 2026-09-08 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.08189v1 Dynamic layer routing reduces the inference cost of Large Language Models (LLMs) by learning to skip layers for individual tokens. Existing methods, however, treat each routing decision as a local operation conditioned solely on the current hidden state which is a formulation that overlooks the sequential, path-dependent nature of routing across depth: earlier decisions shape the representations seen by downstream routers, and the layer-usage objective couples all decisions jointly. We propose History-Aware Routing (HeRo), a dynamic routing framework that resolves this mismatch by introducing a router memory mechanism to maintain an explicit routing state across model depth. The memory is constructed via linear attention, incrementally aggregating preceding routing scores and their induced residual updates into a compact history representation. At each routed layer, the router conditions jointly on this accumulated state and the current hidden representation to select the executed branch. Instantiated for token-wise FFN routing, HeRo trains only lightweight routers and adapters on a frozen backbone, requiring no modification to pretrained parameters. Across Llama 3.1-8B, Llama 2-7B, and Llama 2-13B, HeRo consistently achieves the highest aggregate performance retention among ten baselines. On Llama 3.1-8B, it bypasses 26.87% of model parameters while achieving 100.24% of dense model performance across seven benchmarks, and retains 97.01% while bypassing 38.82% of model parameters under a tighter computation budget. Ablation studies confirm that removing routing history consistently degrades performance, most notably on multistep reasoning and code generation, validating that explicit routing memory enables more accurate and adaptive dynamic routing than solely conditioning on hidden state. --- ### Bridging the Semantic-Utility Gap in Multimodal RAG via Generator-in-the-Loop Alignment Authors: Zhan-Lun Chang, Dong-Jun Han, Seyyedali Hosseinalipour, Mung Chiang, Christopher G. Brinton Published: 2026-09-08 Categories: cs.AI, cs.IR arXiv: https://arxiv.org/abs/2609.08188v1 Vision-language models (VLMs) augmented with retrieval-augmented generation (RAG) benefit from access to external evidence. However, standard retrievers and rerankers optimize for semantic similarity rather than answer utility, creating a preference gap: documents that appear relevant may not help the generator produce a correct answer. Motivated by this, we propose a two-stage generator-in-the-loop alignment framework that closes this gap without human document-level relevance annotations. Our framework consists of two stages: in Stage 1, a VLM generates a hypothetical text passage from the image-query pair, which is used as the retrieval query for dense text search, bridging the image-to-text modality gap. In Stage 2, a cross-encoder reranker adapted with low-rank adaptation (LoRA) is fine-tuned using answer-supervised preference pairs mined from the frozen VLM: given the dataset answer label, a candidate document is labeled positive if the VLM produces the correct answer when given that document as context, and negative otherwise. This generator-guided signal is compatible with multiple alignment loss functions, including contrastive (triplet) loss, pairwise direct preference optimization (DPO), and supervised fine-tuning (SFT), and supports periodic re-mining to refresh preference pairs as the reranker improves. Experiments on VQA-X and A-OKVQA with Qwen3.5-2B and Qwen3-VL-4B-Instruct show that our proposed framework consistently outperforms rank-order, random, and REPLUG-style likelihood baselines under various alignment losses and pool size settings, suggesting that answer-level generator feedback is an effective supervision signal for preference alignment. --- ### Does Deeper Reasoning Compromise Alignment? Revealing and Mitigating of Alignment Collapse in Large Reasoning Models Authors: Yu-Hang Wu, Yu-Jie Xiong, Henghua Zhang, Bairui Zhang, Jia-Chen Zhang, Shaohua Li Published: 2026-09-08 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.08186v1 The emergence of Chain-of-Thought (CoT) has established a robust foundation for Large Reasoning Models (LRMs). While deep reasoning is widely believed to enhance safety alignment, the stability of alignment mechanisms under extended reasoning remains underexplored. This paper challenges the prevailing view by revealing a critical vulnerability: Deep Reasoning May Induce Alignment Collapse. To rigorously quantify this phenomenon, we propose the Alignment Loss Rate (ALR) metric. Our experiments demonstrate that as reasoning depth increases, ALR rises significantly, indicating a severe degradation in model robustness against external perturbations. Capitalizing on this instability, a novel jailbreaking paradigm, Reasoning Trap (RT), is proposed. RT induces the model into extended reasoning to amplify the impact of adversarial attacks, leading to a sharp decline in safety capabilities. To elucidate the mechanism behind this collapse, we identify Attention Dilution as the root cause, arising from the competition for attention between the extended reasoning process and the original input. To mitigate this, Reasoning Residual Alignment (RRA), a lightweight defense strategy that dynamically re-emphasizes the input via residual connections integrated with the reasoning process. --- ### NeoHorse-1: Towards Recursive Self-Improvement via Agentic Post-Training with Routing Harness Authors: NeoHorse Team, Guoliang Cao, Guohao Dai, Tianyu Guo, Kai Han, Hailin Hu, Zihan Jiang, Xiang Kuang, Boxun Li, Yulong Li, Zehua Pei, Yuchuan Tian et al. Published: 2026-09-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.08183v1 Recursive self-improvement (RSI) requires a concrete mechanism through which an AI system observes its capabilities and converts that evidence into the next round of learning. We present NeoHorse-1, a family of agent-native models developed to explore this path through agentic post-training. Our system combines a heterogeneous model pool with intelligent routing, recording the predicted capability demand, selected service tier, and subsequent interaction for each user turn. These records are converted into training examples that preserve interleaved reasoning, tool calls, and harness context, and are admitted through structural validation, six-dimensional semantic evaluation, and subscene-level labeling. Routing signals organize supervised fine-tuning into a three-stage curriculum and extend to routing-guided on-policy distillation, where a teacher supervises student-generated responses under the same progression. Capability-guided allocation then converts evaluation feedback into the next training mixture, closing an evaluation-selection-update loop in which what the system learns to do shapes what it learns from next. Across eleven benchmarks covering harness-based agents, tool use, coding, and instruction following, post-training raises the macro-average from 58.94 to 64.87 at 4B and from 65.60 to 69.04 at 9B, substantially narrowing the aggregate gap between the post-trained 4B model and the 9B base model. NeoHorse-1 provides an initial prototype of this feedback-driven process and a path toward harness-mediated RSI across successive iterations. --- ### WSPolypNet: Weakly Supervised Polyp Localization in Colonoscopy Videos Authors: Giseong Hwang, Minjae Jo, Yeonghyeon Park, Kyeonghun Kim, Seoyeon Han, Donghoon Han, Haneul Kim, Yului Jeong, Insung Hwang, Pa Hong, Ken Ying-Kai Liao, Nam-Joon Kim Published: 2026-09-08 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2609.08182v1 Because dense frame-level annotation of colonoscopy videos is costly, we propose WSPolypNet, a weakly supervised framework for polyp localization using only video-level labels. WSPolypNet employs a 3D convolutional neural network trained with video-level supervision to generate class activation maps (CAMs), which identify candidate polyp regions without requiring frame-level spatial annotations. The CAM-derived localization cues are further enhanced using a multi-view strategy and provided to MedSAM2 as point prompts. MedSAM2 then propagates segmentation masks across the video, refining the coarse localization cues according to polyp boundaries. WSPolypNet achieved CorLoc scores of 47.80%, 43.68%, and 35.01% at IoU thresholds of 0.3, 0.5, and 0.7, respectively, compared with 36.87%, 33.72%, and 27.94% in the single-view setting. For small polyps, the multi-view strategy improved CorLoc@0.5 from 16.01% to 30.97%. The framework also achieved a recall of 94.51%. These results demonstrate the potential of weakly supervised spatiotemporal learning to substantially reduce spatial annotation requirements for polyp localization in colonoscopy videos. --- ### Less Is Personal: Learning Minimal Sufficient User Profiles for Personalized Language Models Authors: Minghang Liu, Qiang Qiu, Yuanzhuo Wang, Huawei Shen, Xueqi Cheng Published: 2026-09-08 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.08180v1 Retrieval-augmented personalization enables large language models to produce more accurate and preference-aligned outputs using relevant records retrieved from user histories. Personalized language models typically prepend a fixed number of retrieved user records, even when additional history is redundant, harmful, or unrelated to a user's distinctive behavior. We study minimal sufficient personalization: constructing the least costly ordered profile for each input while preserving the utility achievable from a retrieved candidate pool. We introduce ENOUGH, a method that iteratively appends behavioral records or emits STOP to construct profiles with adaptive lengths. Offline, bounded counterfactual search evaluates profile prefixes by jointly considering downstream gains, user specificity, and token costs. The resulting long-horizon targets are distilled into a multi-head value controller with explicit ranking and stopping supervision. At inference, the controller selects and orders records through lightweight decisions, and the frozen generator is invoked once after stopping. Extensive experiments on six personalized tasks demonstrate that ENOUGH consistently outperforms strong heuristic and retrieval-augmented baselines in both effectiveness and efficiency, achieving minimal sufficient profiles that preserve personalization utility while reducing unnecessary context costs. --- ### A Theory of Reliable Self-Evolution for Agent Harnesses Authors: Qianshu Cai, Yonggang Zhang, Jun Nie, Maohao Ran, Huajiang Zheng, Jun Song, Xinmei Tian, Yike Guo, Wei Xue Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08175v2 In harness self-evolution, agents modify their own prompts, code, tools, and orchestration while keeping the underlying language model fixed. Recent work has shown that agents can improve themselves in response to task failures and achieve substantial performance gains. However, gains on failed tasks do not automatically ensure that performance on previously successful tasks is preserved, raising concerns about reliable adoption. In this work, we provide a theoretically grounded condition under which a self-evolved harness can be reliably adopted. We then propose a validation rule to make the evolved system satisfy the reliable adoption condition with theoretical guarantees. Consequently, the system can achieve progressive improvement through evolution. This leads to a natural question: Does reliable self-evolution have a performance ceiling, and which factors govern this ceiling? Our theoretical results show that the performance ceiling is determined by the costs of verification and evaluation. On the other hand, self-evolution may stall in practice. In this regard, we show that experimental evidence on agent performance shifts can be used to identify the sources of stagnation. Our work thus establishes a theoretical framework for understanding and advancing reliable harness self-evolution. --- ### OntologyBench: Can Dense Retrieval Satisfy Structured Biomedical Constraints? Authors: Xiao Yu Cindy Zhang, Wyeth Wasserman, Jian Zhu Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08174v1 We introduce OntologyBench, a tiered biomedical retrieval benchmark comprising 471,854 training and 125,744 evaluation query-document relevance pairs across concept grounding, relational retrieval, and compositional phenotype-based retrieval. Although these tasks can be tractable using ontology-aware reference methods, across task tiers, embedding performance is generally lower on relational and compositional tasks than on concept-grounding tasks. Fine-tuning on ontology-derived supervision improves performance on several relational and compositional tasks, whereas the evaluated reranking and LLM-based candidate-scoring methods provide little or no end-to-end improvement. Errors frequently reflect diseases matching only subsets of the phenotype evidence. These findings indicate that the evaluated embedding and reranking configurations do not reliably recover the compatibility encoded by the selected ontology relations and phenotype combinations and motivate retrieval systems that better integrate learned representations with structured biomedical knowledge. --- ### Key Path Identification for Resolving Knowledge Conflicts via SAE-based Steering Authors: Wenbo Zhang, Zhongxiang Sun, Zhiguang Han, Jun Xu Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08173v1 Sparse autoencoder (SAE)-based steering has been widely used to address knowledge conflicts by guiding LLMs to be more faithful to the contextual knowledge. Existing methods usually perform mass steering, which modifies a large batch of SAE features identified via correlation-based methods. However, due to the inaccurate correlation and the neglected feature interactions, mass steering methods fail to precisely identify the features that play the key roles in steering and introduce a large number of redundant ones, which add noise and weaken the steering effects. Our empirical studies reveal that steering only a small subset of the identified features can achieve comparable or even better performance. Motivated by this finding, we propose Key Path Identification (KPI), a novel method that identifies key steering features characterized by strong causal dependencies with both upstream and downstream features. From these features, KPI constructs key paths and steers through less feature modifications. In this way, KPI advances SAE-based steering from quantity-driven to quality-focused, offering a perspective for more precise and interpretable model editing. Experiments in RAG tasks with knowledge conflicts show that our method improves the accuracy by 18% on average compared to the best baseline of mass steering, effectively filtering redundant features, alleviating side effects and demonstrating the core role of key paths in steering. --- ### Optimal Slice-Adaptive Tuning of Hybrid Slice Sampling Authors: Trevor Campbell Published: 2026-09-08 Categories: stat.CO, cs.LG, stat.ML arXiv: https://arxiv.org/abs/2609.08172v1 Slice sampling is a Markov chain Monte Carlo algorithm that draws its next state uniformly from a "slice"---a super-level set of the target density function---at each iteration, thereby providing automatic local adaptivity to the scale of the target. In practice the exact slice is not known, so general-purpose implementations use an approximate slice that is grown from a starting interval of length $w>0$, with a computational cost that depends on $w$. This work presents an analysis of the average per-iteration number of target density evaluations, as a function of $w$, of hybrid slice sampling with various slice-finding schemes for targets with contiguous slices. The paper uses the results of the analysis to develop automated, slice-adaptive tuning schemes along with suboptimality bounds and asymptotic convergence guarantees. Simulations demonstrate that the tuning schemes reliably yield near-optimal slice-adaptive tuning with essentially no dependence on the initial setting of $w$. --- ### EviSI: An Evidence-Based Evaluation Agent for Simultaneous Interpreting Authors: Ben Yan, Zongyao Li, Xiaoyu Chen, Daimeng Wei, Weidong Liu, Huan Zhao, Chong Li, Yaode Wang, Yuzhe Shang Published: 2026-09-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.08171v2 Low-latency simultaneous speech-to-speech translation must keep pace with ongoing speech while preserving key information. To meet these demands, systems use segmentation, reformulation and condensation to reorganize and rephrase information. However, metrics developed for text translation, including BLEU and COMET, may not consistently distinguish faithful adaptations from semantic errors. We propose EviSI, a large language model evaluation agent combining Multidimensional Quality Metrics (MQM) with criteria developed with professional interpreters. Shared source evidence guides assessment across four dimensions: Anchor, Event, Logic and Fluency. Verified errors are deduplicated before deterministic scoring. On human-rated English to Chinese and Chinese to English data, EviSI recovers the aggregate English to Chinese human system ranking. Mean within-dataset Kendall correlations for system rankings reach 0.707 and 0.467, respectively, exceeding evaluated BLEU and COMET baselines. A multilingual extension to five directions without human ratings retains the dimensions and scoring rule, showing positive system ranking correlations with COMET throughout. --- ### Boundary Voting Network for Ambiguity-Aware Timestamp-Supervised Action Segmentation Authors: Runzhong Zhang, Yueqi Duan, Yang Chen, Weipeng Hu, Chen Cai, Suchen Wang, Yap-Peng Tan Published: 2026-09-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.08167v1 Timestamp-supervised action segmentation aims to segment and classify actions in untrimmed videos with a random frame annotated per action. Precisely localizing action boundaries from timestamp annotations is crucial for this setting, as it enables generating framewise pseudo-labels and applying the well-explored fully-supervised training. However, prevailing methods struggle with intrinsic uncertainty in boundary localization due to less discriminative features in action-transiting regions. This imprecise boundary estimation significantly reduces the stability and reliability of the generated pseudo-labels in ambiguous action-transiting regions, consequently resulting in performance deterioration of the trained segmentation models. In our paper, we introduce the boundary voting network that mitigates feature ambiguity by hierarchically propagating video-level global prior knowledge into local action-transiting regions. By generating key action representations as votes throughout the video and targeting action-transiting regions, all votes collaboratively contribute to action-transiting feature enhancement and boundary localization refinement. Extensive experiments demonstrate the effectiveness of our method on GTEA, 50Salads, and Breakfast datasets. --- ### AI for AI: Optimizing Additional Infrastructure Build-out to Power Artificial Intelligence Data Centers Authors: Alexander Crosier, Kyle Onghai, Ronnie Sircar Published: 2026-09-08 Categories: q-fin.GN, cs.AI, econ.GN arXiv: https://arxiv.org/abs/2609.08166v1 The twenty-first century's transformative technology, artificial intelligence, is increasingly constrained by the twentieth century's transformative technology, the electricity grid. Rapid growth in electricity demand from data centers is leading to higher electricity prices, without a compensating supply-side response. We develop a framework linking data-center load growth, available generation capacity, and market-clearing prices to understand this phenomenon. We first analyze a deterministic model to show how differing estimates of demand and supply growth rates affect prices. We then model the expansion of new data centers and their associated electricity demand, together with build-outs of new electricity supply, as stochastic processes,resulting in probabilistic distributions of supply, demand, and prices rather than a single forecast. Finally, we formulate generation expansion as a stochastic control problem in which a revenue-maximizing investor dynamically chooses the intensity of supply-side investments. The analysis highlights a central challenge of the data-center build-out: even when rapid demand growth increases the need for new generation, the uncertainties related to load forecasts, development execution risks, and value cannibalization from overbuilding capacity may weaken incentives to invest at the pace required to keep electricity prices stable. --- ### A Transformer-Based Delta Expression Encoder for Psilocybin Transcriptional Response: Architecture, Representations, and Biological Validation Authors: Sai Jayakumar Published: 2026-09-08 Categories: q-bio.GN, cs.LG arXiv: https://arxiv.org/abs/2609.08165v1 Understanding why individuals respond differently to psilocybin requires modeling the drug's transcriptional perturbation signature at the cell-type level. I present a Transformer-based delta expression encoder that learns to classify differential gene expression status - upregulated, downregulated, or neutral - from single-nucleus RNA-sequencing data, without supervision from pathway annotations or prior biological knowledge. The model is trained on pseudobulk profiles from 623 examples spanning 18 cell types, 2 drug conditions, and 6 timepoints derived from the Liao et al. 2025 dataset, and achieves 69.4% weighted classification accuracy. Three principal findings are reported, alongside one direct test of a published hypothesis that returned a result inconsistent with that hypothesis. First, per-cell-type classification accuracy ranges from 28.3% (L2/3 IT, a primary HTR2A-expressing psilocybin target) to 99.6% (endothelial cells), consistent with known psilocybin response biology. Second, psilocybin-induced transcriptional downregulation is significantly more stereotyped across individuals than upregulation (Mann-Whitney U=18615.0, p<0.0001), a novel finding with a cortical depth gradient across excitatory subtypes. Third, attention-guided gene co-regulation analysis recovers drug-specific modules without pathway supervision. Separately, a direct test of whether baseline HTR2A expression predicts drug-response separability across cell types found a significant negative correlation (Spearman r = -0.7088, p = 0.0021), the opposite of what a simple HTR2A-gating account would predict. --- ### Dual-Layer Semantic-Spatial Belief Mapping for Aerial Object Goal Navigation Authors: Jianqiang Xiao, Xiang Deng, Yuexuan Sun, Yanjin Wu, Wenbiao Yan, Liqiang Nie Published: 2026-09-08 Categories: cs.RO, cs.AI arXiv: https://arxiv.org/abs/2609.08164v1 Aerial Object Goal Navigation (ObjectNav) requires an unmanned aerial vehicle (UAV) to locate a described target in an unknown outdoor environment using onboard visual observations. Vision-language models (VLMs) can interpret open-ended target descriptions and visual observations, but their frame-level outputs are often noisy, sparse, and spatially transient. We propose AeroBelief, a dual-layer semantic-spatial belief mapping framework that transforms transient VLM observations into persistent spatial guidance. It separates broad contextual plausibility from target-specific evidence: an intuition layer accumulates scene-level semantic cues for exploration, while an evidence layer preserves qualified target-specific observations for approach and confirmation. Evidence-gated fusion combines the two layers into spatial belief hotspots. We further introduce object-conditioned visual reasoning with conservative evidence qualification to improve observation reliability before spatial accumulation. In parallel, egocentric regional guidance converts quadtree coverage into UAV-centered, yaw-aligned directional proposals and stabilizes them through temporal commitment. Its regional scoring is independent of semantic belief values, maintaining exploration pressure and reducing repeated low-gain search. Experiments on the UAV-ON benchmark show that AeroBelief achieves the best reported overall SR, OSR, and SPL among the compared methods, reaching 21.61%, 35.57%, and 10.62, respectively. These results support the effectiveness of persistent semantic-spatial belief, conservative evidence qualification, and temporally stable regional guidance for aerial ObjectNav. --- ### WorldAgen: Unified State-Action Prediction with Test-Time World Model Training Authors: Chi Wan, Kangrui Wang, Yuan Si, Pingyue Zhang, Manling Li Published: 2026-09-08 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.08162v1 How can vision-language-action (VLA) models adapt to new environments where world dynamics shift? While recent research has combined world modeling and action prediction to improve VLA performance, existing methods largely rely on pretraining on static datasets, without mechanisms for active adaptation at deployment time. As a result, these models often fail to generalize when deployed in unseen scenarios with novel object configurations or dynamics. We present WorldAgen, a unified framework that jointly learns world modeling and action prediction while enabling Test-Time Training (TTT) to adapt to new environments. WorldAgen employs a shared Transformer backbone with two heads: (1) a world model head that predicts future states from past state-action trajectories, and (2) an agent model head that predicts actions conditioned on task instructions. We design a Mixed Unidirectional Attention Mask to separate these two models. During test time, WorldAgen samples exploratory actions, collects ground-truth state transitions, and performs lightweight TTT updates to refine its world model. This adaptation improves the model's understanding of the environment and leads to more accurate action predictions. Experiments on the CALVIN and LIBERO benchmarks demonstrate that our baseline model achieves comparable, and in some cases superior, performance to current state-of-the-art approaches. Moreover, with TTT on a small number of samples, our method surpasses existing state-of-the-art models, highlighting the effectiveness of adapting world models at inference time. --- ### Snugi-AI-v2 @ eRisk 2026 Task 2: Early Depression Detection via a Learned Stopping Policy with Sustained Confidence Gate Authors: Yuwen Chiu Published: 2026-09-08 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2609.08161v1 We describe the Snugi-AI-v2 submission to eRisk 2026 Task 2, the second edition of contextualized early depression detection from Reddit discussions. Our central contribution is a learned MLP stopping policy trained to directly optimize ERDE50, replacing the fixed and tiered threshold strategies used in all prior eRisk Task 2 submissions. Combined with a sustained confidence gate that commits only after N=3 consecutive rounds of high policy confidence, the system reduces false positives caused by transient emotional posts without sacrificing recall. The pipeline encodes each discussion thread with a frozen MentalRoBERTa model, maps the accumulated representation to a depression probability via an MLP classifier, and delegates the timing decision to the learned policy. Our best run achieves F1 = 0.73 (Run 1) and F_latency = 0.70 (Runs 0 and 3), with a median alert round of 8 out of 500, completing the full evaluation in 1 hour 26 minutes, the fastest among all complete-submission teams. We report a systematic ablation across five runs spanning two encoder variants, four stopping strategies, and three gate values, along with negative results from GRPO policy training, BDI-II post filtering, MentalLongformer encoding, and DeBERTa ensembling. Code: https://github.com/chiuyuwen91/erisk-2026 --- ### When Metrics Reward the Worst Translations: Internalizing Cultural Reasoning for Social Media Translation Evaluation Authors: Yiwen Qiu, Linjuan Wu, Dingming Li, Yizhou Liu, Zixuan Wang, Haolei Xu, Ye Guo, Daoxin Zhang, Weiming Lu, Yongliang Shen Published: 2026-09-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.08156v1 Automatic translation quality metrics trained on general-domain corpora systematically fail on social media content, where communicative intent is encoded in culturally loaded expressions (internet slang, homophonic ciphers, and platform-specific idioms) rather than surface token patterns. We conduct a systematic empirical analysis demonstrating that standard metrics including COMET, XCOMET, and BERTScore exhibit near-zero or negative correlation with human cultural judgments, and even display a severity inversion in which scores increase as translation quality deteriorates. We further show that this failure extends to large language model judges: Qwen3-235B achieves Cohen's kappa of only 0.162, revealing that the bottleneck is not reasoning capacity but cultural grounding: models lack the domain-specific cultural knowledge needed to identify which aspects of a translation require scrutiny. To address this, we propose CuRIL, a reinforcement learning framework that internalizes cultural reasoning: cultural annotations are prepended inside the model's reasoning, excluded from policy gradients via a token-level loss mask, and injected with a probability that decays to zero over training, progressively forcing autonomous cultural judgment. On a 1,444-sample human-annotated social media translation benchmark, Qwen3-8B trained with CuRIL achieves Cohen's kappa 0.370 and Exact Match accuracy of 45.22%, approaching Gemini-3.1-Pro with 30x fewer parameters and surpassing models up to 235B in scale. We further demonstrate that our judge produces reliable reward signals for downstream translation optimization, reducing the low-quality translation rate by over 20 percentage points under independent human evaluation. --- ### SciFigure2Code: An AI-Reconstructed Benchmark for Scientific Figure-to-Code Authors: Wentao Li, Yibo Wu, Yizhe Chen, Ruixuan Chen, Jiangjie Qiu, Yijun Li, Zhao Leyi, Xiaonan Wang Published: 2026-09-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.08155v1 Scientific figures are the interface through which research claims are inspected and reused, but final published panels rarely expose the data or plotting code that produced them. Recovering this hidden provenance from pixels is therefore underdetermined. We introduce SciFigure2Code, an AI-reconstructed benchmark that instead evaluates presentation recovery: generating editable Python programs that preserve how a scientific panel is arranged and read. Role-specialized Codex agents generate, execute, visually refine, and audit silver-standard presentation programs that capture geometry, visual hierarchy, encodings, annotations, and typography without claiming to recover original measurements or author source code. This reconstruction-and-audit protocol turns final published panels into auditable reference packages; the resulting resource contains 6,740 reviewed panels and SciFigureBench, a balanced 337-panel test set across 31 chart subtypes, five domains, and three complexity levels. Across 14 zero-shot models in image-only and caption-assisted settings, execution, multi-component layouts, axes, legends, and scientific labels remain weak. Claude Opus 4.7 achieves the highest image-only Overall score, Claude Opus 4.6 leads caption-assisted reconstruction, and two-stage plan-then-code prompting improves Overall for all four tested models. SciFigure2Code provides an auditable testbed for agents that construct editable, visually faithful scientific figure presentations. --- ### Geodesic-informed Generative Diffusion Model For Topology-preserved Image Video Generation Authors: Nian Wu, Nivetha Jayakumar, Jiarui Xing, Miaomiao Zhang Published: 2026-09-08 Categories: cs.LG, cs.CV, cs.GR arXiv: https://arxiv.org/abs/2609.08153v1 Generative diffusion models have emerged as a class of powerful techniques for various imaging applications, including but not limited to synthesis, reconstruction, and segmentation. Despite their success, current generative models pose two key limitations. First, they primarily rely on image intensity and texture information, with limited attention to underlying object geometry. As a result, they do not guarantee geometric or topological consistency during the generation process, which is a crucial requirement for high-stakes domains such as computational anatomy, biology, and robotics, where preserving object structure is critical. Second, existing models fail to explicitly learn or represent shape changes in the generative process. Such deformation dynamics remain occluded within network parameters; hence leaving the transformation process uninterpretable and physically uninformed. To address these challenges, we introduce IGG (Image Generation informed by Geodesic dynamics), a novel framework that integrates topology-preserving geodesic principles into the diffusion-based generative process. In contrast to conventional methods that operate in image intensity space, IGG learns and synthesizes diverse samples within geodesic deformation spaces, where geometric object changes are learned as smooth and invertible smooth mappings from a given template/source image. Our code is publicly available at https://github.com/nellie689/IGG. --- ### Topology-induced Operators Reveal Complementary Graph Representations without Training Authors: Meng Qin, Jinqiang Cui, Hongwei Zheng, Weihua Li, Sen Pei Published: 2026-09-08 Categories: cs.LG, cs.SI arXiv: https://arxiv.org/abs/2609.08152v1 Graph representation learning has largely focused on designing increasingly sophisticated models to transform graph topology into vector representations, or embeddings. However, the extent to which embedding quality depends on model learning, rather than on the underlying topological transformations, remains unclear. Here, we show that informative embeddings can be derived without complicated model design and gradient-based training. Propagating random features through implicit hierarchical structures induced by random walks and anonymous walks yields embeddings that capture node proximity and structural role, respectively. These two training-free embeddings preserve complementary aspects of graph organization and perform competitively with classic and recent methods across various node-, edge-, and graph-level tasks. They often require substantially less computation, resulting in a favorable quality-efficiency trade-off. Combining the two types of embeddings further improves inference quality of some tasks compared with using either embedding type alone. Our results suggest that informative graph embeddings can arise from carefully chosen topological transformations before any learning operation is applied. --- ### SWE-Bench Pro Verified: A Reliable Benchmark for Software Engineering Agents Authors: Pujun Zheng, Zixin Shang, Shufan Jiang, Wenhui Tian, Dongsheng Zhu, Zerun Ma, Dingbo Yuan, Qi Zhang Published: 2026-09-08 Categories: cs.AI, cs.SE arXiv: https://arxiv.org/abs/2609.08149v2 SWE-Bench Pro has emerged as a standard benchmark for evaluating software engineering agents on challenging repository-level tasks. However, our analysis work show that its evaluation is undermined by two sources of unreliability: reward hacking, enabled by leakage of gold solutions or hidden evaluation information, and task quality issues, including misleading problem statements and improperly scoped tests. These issues can inflate benchmark performance and obscure agents' true coding ability. We present SWE-Bench Pro Verified, a verified version of SWE-Bench Pro that addresses both problems. Our approach combines anti-hacking safeguards that eliminate major leakage channels without disrupting normal agent functionality, with task refinement that minimally corrects inconsistencies within flawed instances. Evaluations on SWE-Bench Pro Verified reveal that some models perform substantially worse than previously evaluated, suggesting that existing results on SWE-Bench Pro may overestimate real software engineering capability. SWE-Bench Pro Verified offers a more trustworthy benchmark for assessing software engineering agents. --- ### MRI-Guided Reslice-Refined Cross-Slice SDF Reconstruction of the Left Ventricle from Cardiac MRI with Sparse Axial Supervision Authors: Quanxin Zheng, Shuai Zhao Published: 2026-09-08 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.08148v1 Reconstructing a three-dimensional left-ventricular (LV) endocardial surface from cardiac magnetic resonance (CMR) data is challenging when supervision is available on only a small number of axial slices. Through-plane geometry is weakly constrained, and automatically generated two-dimensional masks can propagate segmentation errors into the recovered shape. We present MR-RS-SDFR, a per-case implicit signed distance field (SDF) framework that reconstructs a continuous LV surface from a CMR volume and sparse axial weak masks. The method first builds a cross-slice SDF initialization from axial and longitudinal geometric cues and then refines the field using two complementary signals: MRI edge-field normal alignment, which provides an image-derived boundary cue independent of the weak masks, and differentiable reslice Dice and contour consistency, which preserve agreement with the observed planes. We evaluate three weak-mask generators -- LOO TransUNet, LOO nnU-Net, and an off-the-shelf Medical SAM3 model used without MM-WHS-specific training or fine-tuning -- and five sparsity levels from 4 to 64 axial planes. In the sparse-16 setting, final MR-RS-SDFR reconstruction reaches 0.928 Dice and 3.80mm HD95 with Medical SAM3 masks. The upstream generators do not exhibit a single common ranking across 2D and dense 3D segmentation, and nnU-Net- and Medical-SAM3-driven sparse reconstruction achieve the same mean final Dice despite different upstream error profiles. Across all three sparse-16 mask sources, MR-RS-SDFR is numerically better than protocol-matched full GHD+DVS in both Dice and HD95. Final Dice improves markedly from sparse-4 to sparse-16 and then saturates at the reported precision through sparse-64. These results support MRI-guided per-case SDF refinement as a reconstruction strategy that remains effective across weak-mask generators and supervision densities. --- ### ConversationalVoice: Full-Duplex Speech Data from Real Conversations through Source-Faithful Reconstruction and Conversation-Grounded Expansion Authors: Richard Yucheng He, Baodong Cao, Chen Xu, Yihang Liu, Tairan Chen Published: 2026-09-08 Categories: cs.CL, cs.SD, eess.AS arXiv: https://arxiv.org/abs/2609.08147v1 Full-duplex speech models require training data that preserves turn-taking, overlap, interruption, and backchannel behavior, yet these signals are entangled across speakers in noisy real-world recordings. We present Conversational Voice, a pipeline that converts real two-speaker excerpts into three complementary training-data artifacts. (1) Separation recovers speaker-specific tracks with stable speaker assignments, a canonical transcript, and naturally observed interaction timing. (2) Reconstruction generates speech in matched voices from a fixed source transcript, reconstructs the source turn order, pauses, and overlaps, and adds word-level alignment and delivery instructions. (3) Expansion generates new dialogue constrained by the source context, speakers, and observed interaction pattern. Automatic speaker-verification metrics remain strong across stages, with same-speaker similarity of 0.983-0.991 and positive discrimination margins of 0.199-0.209. Predicted speech quality (NISQA MOS) is 3.56 for separation, 4.41 for reconstruction, and 4.61 for expansion. A Gemini-based automatic evaluator assigns expansion mean scores of 4.94/5 for contextual coherence and 4.80/5 for dialogue naturalness. Expansion and reconstruction exhibit broadly similar interaction profiles; expansion's turn, overlap-event, backchannel, and interruption rates are 4.6%, 8.0%, 13.2%, and 16.0% lower, respectively. We evaluate data properties only; downstream gains in full-duplex model training remain for future work. --- ### IGT @ FinMMEval 2026 Task 2: Question-Type Prompting with Targeted Extraction for Multilingual Financial QA Authors: Yuwen Chiu Published: 2026-09-08 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.08139v1 We present the IGT system for PolyFiQA Task 2 of the FinMMEval Lab at CLEF 2026, a multilingual financial question answering task over English SEC filings and multilingual news articles (English, Chinese, Japanese, Spanish, Greek) for four companies. Our central observation is that the 344 development questions divide into two families requiring fundamentally different approaches: structured numeric types (R&D ratio, cash flow, capital expenditure) are best answered by direct keyword extraction on filing text, while synthesis types (investment strategy, capital allocation, top-three revenue focuses) require rule-based multilingual news passage selection. A dataset analysis reveals that 17-18 of 19 ground-truth reference answers per synthesis type share an exact evidence label prefix, whose unigram tokens contribute directly to ROUGE-1 overlap. The final system achieves development ROUGE-1 approximately 0.395, a 60% relative improvement over a generic RAG baseline (approximately 0.247), and ranks 3rd of 12 teams on the official test set with ROUGE-1 = 0.3071, Precision = 0.2821, and Recall = 0.4044. --- ### GPU-Enabled Large-Scale Optimization Using Randomized Linear Algebra Authors: Pratik Rathore, Zachary Frangella, Parth Nobel, Xuning Hu, Madeleine Udell Published: 2026-09-08 Categories: cs.LG, math.OC, stat.ML arXiv: https://arxiv.org/abs/2609.08136v1 This paper introduces rlaopt, a PyTorch-based package for large-scale optimization and scientific computing using randomized numerical linear algebra (RandNLA). Despite substantial progress in RandNLA-based algorithms, few implementations combine GPU acceleration with a simple interface for specifying optimization problems. rlaopt addresses this gap by providing GPU-enabled solvers for positive-definite linear systems and convex empirical risk minimization with constraints and regularizers. These solvers use RandNLA to accelerate conjugate gradient (NystromPCG), operator splitting (NysADMM), and stochastic gradient methods (SAPPHIRE). Moreover, rlaopt includes a modeling language that lets users specify problems using natural mathematical syntax. rlaopt automatically checks compatibility with the selected solver and performs the required problem decomposition. The solvers also support differentiation through their iterations, enabling applications such as hyperparameter tuning. Experiments on ridge regression, bounded multinomial logistic regression, and bounded elastic net identify when randomized preconditioning improves performance and demonstrate substantial speedups from GPU execution. The package is open-source under an Apache license, with source code at https://github.com/udellgroup/rlaopt and version 0.1.0 available on PyPI. --- ### KBBQ: A Predictive Noise Law and the Limits of Spectrum Flattening in FP4 Quantization Authors: Lexington Whalen, Yuki Ito, Ryo Sakamoto Published: 2026-09-08 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.08135v2 We develop a second-order theory of quantization noise in matrix multiplication in which the quantization format is characterized by the variance it assigns to each element. The constant variance profile of integer quantization recovers existing integer-noise theory, while the multiplicative profile of floating-point rounding reduces the data dependence to a scalar, the participation factor $κ$, yielding a closed-form signal-to-noise-ratio law. The resulting functional also admits a closed-form upper bound $κ^{*}$ that no function-preserving linear transform can exceed and that is attained by a recent state-of-the-art method. Building on this analysis, we introduce KBBQ (\textbf{K}appa-\textbf{B}raked \textbf{B}lockwise \textbf{Q}uantization), which parameterizes the extent to which a transform approaches this ceiling. At W4A4, across four base models and two FP4 formats, KBBQ outperforms the prior state of the art without additional deployment-time computation. --- ### Sparse Data Augmentation for Optimization with Provable Guarantees Authors: Behrooz Tahmasebi, Melanie Weber Published: 2026-09-08 Categories: cs.LG, cs.AI, math.OC arXiv: https://arxiv.org/abs/2609.08133v2 In nonconvex optimization problems arising in geometric machine learning, data augmentation is commonly used to promote invariance by averaging empirical losses over transformations of the data. Computing the fully augmented objective, however, requires access to every element of the transformation group $G$, which may be prohibitively expensive when $G$ is large or accessible only through sampling. We study whether full augmentation can instead be approximated using a small, fixed sample of transformations acquired before optimization and reused thereafter. Under suitable regularity conditions, we show that, with probability at least $1-δ$, gradient descent (GD) on the resulting sparsely augmented objective returns an $\varepsilon$-stationary point of the fully augmented objective using $\mathcal{O}\bigl((\log |G|+\log(1/δ))/\varepsilon^2\bigr)$ group-transformation-oracle queries. By comparison, standard group stochastic gradient descent (group-SGD), which samples a fresh transformation at every iteration, uses $\mathcal{O}(1/\varepsilon^4)$ transformation queries. Therefore, gradient descent with fixed sparse augmentation requires fewer transformation queries than both GD applied to the fully augmented objective and group-SGD. Our proof techniques, which may be of independent interest, establish a uniform approximation of the full group-averaged gradient field by a random group average using spectral properties of group-induced operators and tools from representation theory. --- ### Jacap: Robust KV Cache Eviction via Jacobian-Based Nonlinear Information Capacity Preservation Authors: Jiaming Yang, Chenwei Tang, Liangli Zhen, Chenyang Zhang, Jiancheng Lv Published: 2026-09-08 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.08131v1 Key-value (KV) cache eviction is essential for scaling long-context inference in Large Language Models. However, existing policies predominantly rely on empirical heuristics, lacking a rigorous characterization of token utility under the inherently nonlinear softmax attention mechanism. In this work, we rethink KV cache eviction through the lens of local information geometry, modeling the attention process as a nonlinear Gaussian communication channel. By performing a first-order Taylor expansion of the attention mapping, we derive the Jacobian Information Capacity, a novel objective that explicitly captures query relevance, softmax sensitivity, and structural diversity. Guided by this theory, we introduce Jacap, a capacity-aware eviction method that utilizes softmax-aware importance weighting and statistical leverage scores for subset selection. Extensive experiments across diverse architectures and benchmarks demonstrate that \textsc{Jacap} delivers superior performance in most scenarios, particularly in high-compression regimes. --- ### Observe Before You Alert: Adaptive Driver Alerting with Vision-Language Models Authors: Yuhang Wang, Lingyao Li, Hao Zhou Published: 2026-09-08 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.08130v1 Driver alerting from dashcam video requires sequential decision-making under partial observability: a system must decide not only whether a scene is risky, but also when the evidence is sufficient to warn. Most existing accident anticipation models output a binary risk score, leaving ambiguous scenes to be handled by thresholding. We propose VLAlert, a vision-language alerting framework that casts warning generation as a tri-action policy over SILENT, OBSERVE, and ALERT. The OBSERVE action acts as an internal evidence-gathering decision that delays uncertain warnings and changes the next observation window, creating a lightweight perception-action loop for adaptive alerting. VLAlert uses Qwen3-VL-4B as a safety-evidence generator and pools hidden states from structured belief spans to form compact representations for danger estimation and policy prediction. We evaluate VLAlert on VLAlert-Bench, a unified per-tick benchmark from four real-world dashcam alert datasets, and further test transfer to held-out naturalistic ADAS takeover clips. On VLAlert-Bench validation, VLAlert achieves the highest deployment-oriented utility among tested baselines, with DAUS 0.4878 compared with 0.4752 for Open-BADAS, and improves AUROC, AP_tick, F1_t, and balanced accuracy from 0.610, 0.176, 0.276, and 0.581 to 0.689, 0.195, 0.297, and 0.648, respectively. On 221 held-out ADAS-TO-Critic clips, VLAlert improves R@5s from 74.2% to 88.7% and F1 from 0.585 to 0.686. These results indicate that adaptive observation and safety-focused VLM representations provide measurable gains for driver-facing alert decisions. ---