# Research Papers — 2026-09 · bundle page 186 of 266 > Full-text markdown of papers #9251–9300 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/186 [← Previous bundle](https://aipod.dev/papers/archive/2026-09/page/185.md) [Next bundle →](https://aipod.dev/papers/archive/2026-09/page/187.md) ### Explaining Time Series Forecasting with Horizon-Resolved Attribution Authors: Seunghan Lee, Jun Seo, Jaehoon Lee, Junhyeok Kang, Sangjun Han, Sungdong Yoo, Minjae Kim, Tae Yoon Lim, Dongwan Kang, Hwanil Choi, Soonyoung Lee, Wonbin Ahn Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.12639v1 Recent advances in explaining time series (TS) models have produced methods that identify which past values a prediction depends on. However, most existing methods return a single importance vector, assuming that every predicted step depends on the same past values. In this paper, we show that this assumption does not hold, as different forecast steps depend on different past values. Motivated by this observation, we propose Horizon-Resolved eXplanation (HRX), which adds a horizon axis to the explanation, so that every forecast step receives its own importance map. HRX is a simple yet effective plug-in framework with three components: 1) an estimator that reads these maps out of any differentiable forecaster without modifying the TS backbone, 2) an evaluation protocol that validates the horizon axis by measuring how much a single forecast step changes when the inputs an importance map ranks highest are removed, and 3) a rank criterion that predicts in advance whether the axis is worth resolving on a given TS. We further show that this step-wise dependence is low-dimensional, as the explanations of all steps are built from a few shared maps whose number does not grow with the forecast length. Extensive experiments across various backbones and datasets show that the improvement comes from the horizon axis and holds for estimators of previous explanation methods. Code is available at https://github.com/seunghan96/HRX. --- ### Geometric-to-Semantic Spherical Transfer Learning for Cortical Sulci Labeling Authors: Saeb Tounsi, Joël Chavas, Pietro Gori, Vincent Frouin, Denis Rivière, Jean-François Mangin Published: 2026-09-11 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2609.12627v1 Deep learning on cortical surfaces faces a dilemma: capturing the complex topology of over 60 nomenclature-dependent sulci per hemisphere requires high-capacity models, yet the extreme scarcity of expert annotations ($N=62$ subjects) inevitably causes overfitting. Standard supervised approaches fail to generalize in this data-scarce regime, particularly for variable and small sulci where topological ambiguity is high. To overcome this limitation, we introduce a Geometric-to-Semantic Spherical Transfer Learning framework. First, we leverage massive unlabeled data (UK Biobank, $\approx$30,000 subjects) to pre-train a spherical encoder using a locally-optimized strategy. By relying solely on continuous surface features (curvature and depth), the relevance of this pre-training is confirmed by the model's ability to detect localized and rare topological traits, such as sulcal interruptions. The downstream labeling task, however, introduces extracted sulcal fundi (lines) as an explicit semantic input. To bridge this dimensional domain gap (from purely geometric to semantic) without causing catastrophic forgetting, these anatomical lines are integrated into the pre-trained backbone via a soft-initialized Topological Prior Injector. Our experiments demonstrate that this approach outperforms fully supervised baselines trained from scratch, achieving a mean Dice of 0.77. Crucially, a local analysis reveals that the self-supervised geometric priors yield the largest performance gains on variable and tertiary sulci (up to 14.8%), confirming that learning the cortex shape is highly beneficial for identifying its rarest parts. --- ### SteerDuplex: Steerable Duplex Speech Dialogue Models Authors: Utkarsh Tyagi, Ramaneswaran Selvakumar, Advait Gosai, Sonal Kumar, Nikhil Barhate, Isabell Sagar, Steven Li, Miheer Bavare, Daniel Quigley, Fabiola Tapia Carrillo, Jose M Patron E, Diego Macías Gutiérrez et al. Published: 2026-09-11 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.12623v1 Full-duplex spoken dialogue models support low-latency turn taking, interruption handling, and backchanneling, yet a key capability remains underexplored: steerability, the ability to reliably shift conversational behavior along attributes such as tone, persona, speaking rate, and voice style in response to user instructions. We introduce a taxonomy of text- and audio-based steerability that identifies substantial gaps in current full-duplex models. To address this gap, we introduce SteerDuplex, a Moshi-based full-duplex speech model fine-tuned on natural conversations and synthetic dialogues targeting instruction following, vocal delivery, reasoning, and duplex interaction. We further apply two-stage reinforcement learning (RL) with hybrid rewards, combining verifiable interaction checks and judge-based semantic feedback to improve timing and response continuity. To evaluate full-duplex spoken steerability, we introduce SteerBench, a benchmark with 390 spoken prompts and 1,067 human-authored binary audio and text rubrics spanning tone, persona, style/accent, and speed/length. On SteerBench, supervised training improves audio-steering average pass rate by 44.5 percentage points over the strongest evaluated open baseline. On Audio MultiChallenge, task average pass rate improves by 7 points over its strongest evaluated open baseline. RL further raises source-clean interruption response from 72.5% to 82.5% and reduces synthetic pause barge-in from 26.5% to 9%. Steering and aggregate task scores remain comparable or higher, while reward probes reveal reward hacking through incomplete responses. Our model and benchmark support systematic research on spoken steerability, with reward analysis showing why timing gains must be evaluated alongside response completeness. --- ### RA-SOD: Reliability-Aware RGB-T Salient Object Detection under Modality Degradation Authors: Hongbo Gao, Zhengyu Li, Xueru Nie, Dihao Zhu, Lijun Zhao, Yunke Wang, Chang Xu Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12622v1 RGB-Thermal (RGB-T) salient object detection leverages complementary cues from visible and thermal modalities to improve robustness in challenging environments. However, in real-world scenarios, the reliability of each modality is inherently unstable: RGB images degrade under low illumination, motion blur, and noise, while thermal imagery often suffers from contrast compression and sensor artifacts. Such degradation introduces unreliable perceptual evidence that can mislead cross-modal fusion and significantly deteriorate detection performance. To address this challenge, we propose RA-SOD, a reliability-aware RGB-T salient object detection framework that explicitly models modality reliability and integrates it into feature learning and cross-modal fusion. First, we introduce a reliability-conditioned representation that adaptively compensates degraded modality features while preserving structural cues. Second, an uncertainty-guided dual-stream refinement strategy progressively corrects cross-modal representations while suppressing unreliable evidence. Finally, we propose a pixel-wise modality competition mechanism that dynamically selects modality cues according to spatial reliability for fine-grained fusion. Extensive experiments on four benchmarks (VT821, VT1000, VT5000, and VT-IMAG) demonstrate that RA-SOD achieves state-of-the-art performance and exhibits strong robustness under severe modality degradation. Code and models are available at https://github.com/zaoxienian/RA-SOD. --- ### Correlation-Guided Fast Machine Unlearning via Hessian Analysis Authors: Ayushi Thakur, Ruchir Gupta, Amit Kumar Jaiswal, Prayag Tiwari Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.12620v1 The increasing adoption of machine learning in network and distributed security systems has created an urgent need for mechanisms that can selectively and efficiently remove the influence of specific training data to eliminate compromised or adversarial data points from production models. Privacy regulations such as GDPR's \emph{right to be forgotten} also pose similar requirements. However, existing approximate unlearning techniques remain computationally prohibitive for deployment in real-world security systems, as they require repeated expensive Hessian-inverse-vector computations for each data point removal, creating a bottleneck when processing multiple related requests in scenarios such as intrusion detection systems, spam filters, and threat intelligence platforms. Thus, we introduce a computationally efficient unlearning framework that identifies correlated data points in the training set and applies a theoretically derived closed-form parameter update rule, achieving an $82\times$ wall-clock speedup over standard influence function unlearning while preserving model utility with a $10^{-2}$ improvement in accuracy over state-of-the-art baselines. Our method establishes theoretical guarantees and ensures numerical stability through Hessian damping. Our evaluation across seven diverse dataset architecture combinations, including large-scale CIFAR-100 with ResNet-50, demonstrates superior forgetting effectiveness, with membership inference attack success rates of 0.660 and tug-of-war scores of 0.950. --- ### ResoSeg: Resonance Tagger using Transformer and Segment Model Authors: Chunkai Li, Junhao Yin, Ke Li, Jingde Chen Published: 2026-09-11 Categories: hep-ex, cs.AI arXiv: https://arxiv.org/abs/2609.12610v1 Deep learning has been widely applied across many areas of experimental high-energy physics, yet existing models address only event-level classification or object tagging and therefore still require reconstruction algorithms tailored to each decay channel. We present the first application of segmentation to resonance tagging at BESIII and introduce ResoSeg, a deep learning model that jointly performs particle-level segmentation and event-level classification, enabling a one-pass analysis of resonance to anything decays while precisely reconstructing the relevant resonance properties. We demonstrate the reconstruction of $η_c$ with $e^+e^-\toπ^+π^-h_c$, $h_c\toγη_c$, $η_c\to\text{anything}$. The model is trained on BESIII-$η_c$ dataset, which is constructed with per-track true labels obtained via a Truth-Matching Algorithm. Experimental results show that the average combined efficiency of ResoSeg is more than double that of the conventional 16-channel approach across energy points from 4.19 to 4.60\,GeV. The model generalizes to unseen energy points, adapts to other $η_c$ production modes through transfer learning, and remains robust against variations in the $η_c$ mass, width, and branching fractions, providing a general, resonance-aware model applicable beyond $η_c$ and BESIII. The source code is available at https://github.com/oashen/ResoSeg. --- ### Direct Preference Density Alignment for Conversational Audio Equalization Authors: Ioannis Stylianou, Sven Ewan Shepstone, Jon Francombe, Pablo Martinez Nuevo, Zheng-Hua Tan Published: 2026-09-11 Categories: cs.SD, cs.AI arXiv: https://arxiv.org/abs/2609.12607v1 Large Language Model alignment typically relies on learned proxy reward models, which significantly increase the memory footprint during training and are notoriously prone to instability and reward hacking. While offline methods like Direct Preference Optimization (DPO) bypass the reward model, they lose the ability to perform online exploration. If no optimization constraints are applied, this can lead to format collapse in bounded, continuous spaces. To resolve this, we propose Direct Preference Density Alignment: An alternative framework that removes the need for a learned proxy reward model while strictly preserving the benefits of online reinforcement learning. We leverage large-scale user data (approximately 90,000 samples) to construct non-parametric preference density maps, establishing an empirical reward surface. In addition to removing the reward model, Direct Preference Density Alignment enables the combination of the online structural grounding of Group Relative Policy Optimization (GRPO) with the targeted offline refinement of DPO. We show that this GRPO+DPO combination achieves the highest performance, and in a blind audio equalization listening test, enables a 1.5B-parameter model to achieve perceptual parity with a carefully prompt-engineered GPT-4o mini baseline, using only a fraction of the inference compute. --- ### Beyond Generation and Accuracy: Diagnosing and Enhancing Visual Chain-of-Thought for Geometry Problem Solving Authors: Zhitong Dong, Jicai Pan, Yingguo Gao, Jingting Ding, Hao Chen, Jinjie Gu Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.12606v2 While multimodal reasoning has advanced rapidly, solving complex geometry problems critically hinges on active visual assistance, such as constructing auxiliary lines, spurring the rise of Visual Chain-of-Thought (VCoT). However, existing evaluations typically assess visual generation quality and final answer accuracy in isolation, failing to examine whether intermediate visual aids are geometrically valid, effectively utilized in subsequent reasoning, or causally responsible for task success. To bridge this gap, we introduce GeoVAD-Bench, a diagnostic benchmark that pairs a fine-grained five-dimensional trajectory diagnosis covering perception, auxiliary quality, utilization, deductive reasoning, and final correctness with controlled No-Aux, Auto-Aux, and GT-Aux intervention settings to systematically isolate intermediate error modes, the causal gains of visual aids, and the resulting autonomy gap. Our findings reveal that while high-quality auxiliary aids offer substantial theoretical gains for geometric problem solving, autonomous generation is frequently hampered by compounding errors across geometric perception, faithful visual manipulation, visual-state grounding, and deductive reasoning. Guided by these diagnostic insights, we establish a specialized data construction pipeline encompassing geometric perception, diagram editing, and interleaved visual-textual reasoning trajectories, and develop a progressive SFT and multimodal RL training framework. The resulting model, GeoWeave-8B, outperforms the base model by +25.3% in final geometric accuracy and achieves a +30.4% gain in process average across the four intermediate diagnostic dimensions. --- ### SIMS: Scale-Invariant Merit-Function-Based Scalarization for Multi-Task Learning Authors: Zebin Chen, Fei Xing, Yang Chen, Hua Liu, Andy HF Chow, Yuhua Qian, Yu Zhang Published: 2026-09-11 Categories: cs.LG, cs.AI, math.OC arXiv: https://arxiv.org/abs/2609.12599v1 Multi-task learning (MTL) requires navigating unavoidable trade-offs among competing objectives. This paradigm is frequently formulated as multi-objective optimization (MOO), where the scalarization is favored to reduce an MOO problem to a single objective. We empirically find that existing merit-function-based scalarization approaches are sensitive to the relative scales of different objectives in practical MTL, where task losses commonly differ by orders of magnitude. The optimization process often favors objectives with larger scales even though the underlying Pareto optimal solutions remains invariant to rescaling (i.e., multiplying an objective by a positive constant). To address this issue, we propose Scale-Invariant Merit-function-based Scalarization (SIMS) for MTL. Specifically, SIMS adopts a transformation-induced merit function to convert the MOO problem of MTL to a single objective that renders optimization invariant to the magnitudes of losses. Theoretically, we prove that the requirement for scale invariance uniquely determines this transformation to be logarithmic. We further show that this general transformation-induced merit function preserves weak Pareto optimality and admits a smooth surrogate with controllable approximation error. Extensive experiments on representative multi-task benchmarks demonstrate that SIMS consistently outperforms existing scalarization methods and achieves state-of-the-art performance. --- ### Poisson-Corrector Complexity Bounds for Moreau--Yosida Unadjusted Langevin Sampling Authors: Yuchen Xin, Zhihua Zhang Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.12594v1 We study the classical Moreau--Yosida unadjusted Langevin algorithm (MYULA) for $π(\,\mathrm{d} x)\propto e^{-f(x)-g(x)}\,\mathrm{d} x$, where $f\in C^2(\mathbb{R}^d)$ is $m$-strongly convex with $L_f$-Lipschitz gradient and $g:\mathbb{R}^d\to\mathbb{R}$ is convex and globally $G$-Lipschitz. For the Moreau-smoothed target $π_λ$ and the MYULA invariant law $\widehatπ_{λ,h}$, we prove \( \sqrt m\,W_2(π_λ,\widehatπ_{λ,h}) =O(h)+\widetilde O(h^{3/4}) \) under $0<h(L_f+λ^{-1})\le c$, with only logarithmic dependence on $λ^{-1}$ in the error coefficients. Combining this estimate with the Moreau approximation bias yields $\widetilde O(\varepsilon^{-4/3})$ iterations to achieve $\sqrt m\,W_2(μ_N,π)\le\varepsilon$, for fixed model parameters and initialization. The proof combines a discrete Poisson corrector with active-trace estimates and a shared-noise bound for the exact--Euler two-point curvature. --- ### Where Decoder Cosine Similarity Fails for SAE Feature Flow Discovery Authors: Hendrik Droste, Christian Medeiros Adriano, Kathrin Korte, Holger Giese Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.12591v1 Foundation models are increasingly adapted through fine-tuning, model editing, and alignment procedures while retaining previously acquired capabilities. Understanding the internal computations that support these adaptations is therefore becoming increasingly important for continual model evolution. Sparse autoencoders (SAEs) provide interpretable feature dictionaries for residual-stream activations and sublayer outputs, but it remains unclear how state features and update features interact to produce downstream residual features. In this work, we focus on MLP updates as a first test case. We construct a transition atlas of triples $s_k + u_j \rightarrow t_\ell$, where a residual-state feature and an MLP-update feature jointly predict a target residual feature, and validate candidate triples by ablating the decoded update feature. In a 20M-token Pythia-160M $L_7 \rightarrow L_8$ run, we find 38,125 strong ablation-effect transitions, but 88.0% have both state-target and update-target decoder cosine similarity below 0.7. As a preliminary cross-model check, a run of 20M-token Gemma-3-4B $L_{21} \rightarrow L_{22}$ causally validates only the top 30,000 ranked candidate triples by ablating the decoded update feature, and 53.6% of strong-effect triples have both state-target and update-target decoder cosine similarity below 0.7. The Gemma result is directionally consistent with Pythia, but weaker, since update-target cosine recovers many of the strongest Gemma effects and the run is not a full-atlas causal validation. Ultimately, our results suggest that feature flow atlases can serve as diagnostics of representation-update mechanisms and thereby inform tools for steering model updates. Future work will validate more complex patterns across layers, models, and SAE families. --- ### Tight Sampling Complexity with Stochastic Gradient Oracles in Fixed Dimensions Authors: Weiming Ou, Xiao Wang Published: 2026-09-11 Categories: math.ST, cs.LG, math.PR arXiv: https://arxiv.org/abs/2609.12590v2 We establish the stochastic-gradient query complexity of sampling smooth strongly log-concave distributions in every fixed dimension $d\geq1$. For $μ$-strongly convex, $L$-smooth potentials with unknown minimizers $x_f^\star$ in the ball $\mathbb{B}(0,μ^{-1/2})$, and unbiased gradient oracles with variance at most $σ^2$, the minimax worst-case expected query complexity is $Θ\left(\log(1+κ)+\frac{σ^2}{μ\varepsilon}\right)$, jointly optimal for the condition number $κ:=L/μ$, variance $σ^2\ge 0$, and TV accuracy $0<\varepsilon\leq 1/10$. In the noiseless setting $σ=0$, this tight complexity $Θ\left(\log(1+κ)\right)$ is independent of $\varepsilon$. Moreover, a gradient-only sampler generates an exact sample with ${O}(\log(1+κ))$ worst-case expected queries. Without previous ball $\mathbb{B}(0,μ^{-1/2})$, exact sampling from any initial point $x_0$ can be implemented with expected cost ${O}\left(\log(1+κ)+\log(1+\sqrtμ\|x_0-x_f^\star\|)\right)$, without knowing the initial distance. We also show that dependence on initial distance is generally unavoidable. --- ### Reproducing and Evaluating the Generalizability of Subliminal Learning in Open-Weight Models Authors: Daan van der Weijden, Nathan Brack, Selene Baez Santamaria Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.12586v1 In this reproduction paper we investigate subliminal learning, a consequence of distillation where teacher models transmit behavioral preference traits through semantically unrelated data. The original paper explores two types of traits (animal preferences and misalignment), three data modalities (number sequences, code, and chain of thought), and several model families. We reproduce their experiments and extend the setup along three axes: new preference categories (actors and politicians), a new task (chess move generation), and an additional open-weight model (Ministral8B). We also run a controlled ablation on the numbers task's answer-space size (1-, 2-, and 3-digit sequences). We focus on open-weight models with accessible checkpoints on HuggingFace, since the original paper's GPT-4.x fine-tuning is no longer available. Our reproduction supports the original paper's claims, but our extensions show they are not universal as transmission strength varies across traits and tasks, and one model shows almost no effect at all. --- ### Clustering-Based Balanced Sampling and Allocation with Data Parallelism for High-Performance Fine-Tuning Authors: Hyunjin Kim, Youngeun Nam, Jaemin Han, Wonhyeok Choi, Jae-Gil Lee Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.12584v1 Instruction-tuning datasets for large language models (LLMs) are often large, redundant, and imbalanced, limiting efficient adaptation. Naive large-batch fine-tuning repeatedly includes overrepresented sample groups while weakly covering underrepresented but informative ones, especially under data parallelism (DP) across multiple GPUs. We propose CluSTER, a Cluster-aware balanced Sampling framework for Training Efficient data Reduction in DP instruction tuning. CluSTER curates a representative reduced dataset through gradient-space clustering and DP-aware balanced allocation, ensuring dual-level coverage across clusters and workers, while preserving the original data distribution by weighted update. As a result, CluSTER reduces redundant computation and improves training stability without compromising model quality. Across multiple instruction-tuning datasets, CluSTER reduces training time by up to 69.6% with almost no accuracy loss compared to prior sampling and data reduction methods. Code is available at https://github.com/kaist-dmlab/CluSTER. --- ### SCOPE-OPSD: Fisher-Conditioned Privileged Subspaces for On-Policy Self-Distillation Authors: Yunmeng Chen, Kunyu Wang, Peihan Li, Yi Wang, Shuyin Xia, Yi Liu, Xinyong Cheng, Dehui Wang, Xiangyong Zhai, Yanxing Liu, Song Liu Published: 2026-09-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2609.12579v1 On-policy self-distillation (OPSD) scores student-generated prefixes with a solution-conditioned self-teacher, yet transfers supervision only through next-token probabilities. We ask whether the aligned final-layer discrepancy offers a useful second channel, and how to test that channel without confusing its geometry with auxiliary strength. SCOPE-OPSD projects the privileged teacher-student residual onto a frozen rank-64 factor estimated from residual covariance and language-model-head Fisher sensitivity. It reuses the forwards already required by OPSD and adds neither rollouts nor inference-time modules. A matched Random control preserves the structured factor's rank and nonzero spectrum and uses per-arm gradient-RMS calibration, isolating the effect of the data-dependent orientation. Across the complete 25/50/75/100-step trajectories for Qwen3-1.7B, 4B, and 8B, Structured is never below Pure OPSD, with strict gains in 11 of the 12 model-checkpoint combinations and an exact tie at 4B step 25. Structured also exceeds matched Random in 10 of the 12 combinations. At step 75 on Qwen3-1.7B, Structured exceeds matched Random by 1.39 Macro Avg@12 points in each of two independent training reruns. A cross-fitted diagnostic also shows 4.40 times greater held-out privileged-gap capture than the matched random orientation. The results support a compact, Fisher-conditioned privileged subspace for short-budget OPSD. --- ### From Collaboration to Capability: Internalizing Routed LLM Experts into Compact Reasoners Authors: Frank Nie, Shuyao Wang, Ethan B. Liu Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.12578v1 A compact controller can coordinate stronger experts by selecting whom to consult, formulating requests, and integrating their responses. We study whether learning from both the controller's decisions and the experts' reasoning and code improves its generation after expert removal. We introduce \textsc{Rivet} for \emph{collaboration internalization}: expert-augmented reinforcement learning applies a shared outcome signal to controller decisions and returned expert spans, and verified trajectory internalization consolidates complete successful interactions through format-aware supervised training. The deployed controller generates reasoning, code, and interaction structure with local Python execution and no external LLM. Across seven competition-mathematics benchmarks, RIVET-1.7B and RIVET-4B achieve average accuracies of $28.25\%$ and $44.16\%$; Stage~II improves RIVET-4B's accuracy after expert removal by $6.49$ points, and GPQA-Diamond results provide evidence of generalization to scientific reasoning. Ablations show gains from ordinary trajectory supervision and additional format weighting, supporting the effectiveness of training on the content and structure of verified collaborations. --- ### SCORE: SubDistribution-aware Collaborative Knowledge Reinforcing for Cloth-Hybrid Lifelong Person Re-Identification Authors: Kunlun Xu, Liangyu Ma, Jiangmeng Li, Xin Tong, Xiaode Liu, Yufei Guo, Jiahuan Zhou Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12577v1 Lifelong Person Re-Identification (LReID) aims to train a unified person retrieval model from a non-stationary data stream. Existing LReID methods mainly focus on scenarios where the clothing of each person is consistent. Recently, the Cloth-Hybrid LReID (CH-LReID) where cloth-consistent and cloth-changing data alternately occur, has emerged as a more practical and challenging scenario. Due to the conflict between clothing-relevant and clothing-irrelevant knowledge, the well-known catastrophic forgetting problem is significantly exacerbated in this task. To address this issue, we propose a SubDistribution-aware COllaborative Knowledge REinforcing (SCORE) framework, where our key idea is explicitly modeling the intra-identity diversity to continually consolidate distinct cloth-consistent and cloth-changing knowledge. Specifically, an Adaptive SubDistribution Modeling mechanism is developed, where a set of distributional subprototypes is assigned to each identity to capture the intra-identity diversity, improving the compatibility between cloth-consistent and cloth-changing knowledge. Then, a Distributional Knowledge Reinforcement scheme is introduced, where the knowledge of old distributional subprototypes is retained in the new ones by a collaborative aligning mechanism. Extensive experiments show that our SCORE achieves the state-of-the-art performance. Our code is available at https://github.com/zhoujiahuan1991/ECCV2026-SCORE --- ### Calibrated Ambiguity in Multimodal Language Models: Humans reach for cultural references, while models describe the picture Authors: Cody Kommers, Mingrui Ye, Evelyn Gius, Daniela Mihai, Hoyt Long, Zheng Yuan, Drew Hemment Published: 2026-09-11 Categories: cs.CL, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2609.12575v1 Ambiguity is often treated as a bug for AI systems to resolve---but in human communication and culture, ambiguity can also be a generative resource. From humour to politics to art, people express themselves in words and images that are open enough to invite different interpretations, yet constrained enough to be interpretable. We operationalise this notion of calibrated ambiguity with a task drawn from the parlour game Dixit. We compare differences in clues generated by human vs multimodal language models, based on a novel coding rubric for calibrated ambiguity, and find that models consistently exhibit ambiguity collapse (i.e., their outputs are over-specified, leaving no room for multiple legitimate interpretations). Unlike human clues, AI-generated clues also exhibit cultural flattening; they almost never make reference to culturally-situated knowledge, even when prompted to use allusion and figurative language. --- ### An Ultra-Widefield Swept-Source OCTA Dataset and a Polar-Gated Mamba Network for Retinal Vessel Segmentation Authors: Yang Liu, Yibing Shen, Keming Zhao, Cenk Jiang, Zhenghang Qian, Zhicheng Du, Chen Xiong, Qidong Shao, Zijun Lin, Yunqi Hu, Jingjing Zhou, Lian Zhang et al. Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12574v1 Ultra-widefield (UWF) swept-source optical coherence tomography angiography (SS-OCTA) enables large-area retinal vascular imaging, yet vessel segmentation at this scale lacks dedicated public benchmarks and comprehensive evaluation for quantitative vascular analysis. We introduce WOIVES, to our knowledge the first publicly available UWF SS-OCTA vessel-segmentation dataset, comprising 206 eyes from 152 participants with a 24x20mm^2 field of view. WOIVES spans emmetropia to high myopia and provides soft probability vessel annotations. We further propose PG-Mamba, a visual state space model that enhances conventional directional scans with two complementary polar-coordinate scan orders. An auxiliary Dynamic FOV Gating module performs spatial modulation at the bottleneck. PG-Mamba outperformed seven competitive approaches on broad segmentation metrics under cross-validation. It achieved the lowest median absolute errors for vessel density, fractal dimension, and vessel length density. WOIVES is publicly available on Zenodo (DOI: 10.5281/zenodo.21904672), and the PG-Mamba code is available at https://github.com/syb1234567/PG-Mamba. --- ### TokenMapper: A Step Toward Interoperable Speech Token Translation Authors: Tal Kozakov, Tal Rosenwein, Eliya Nachmani Published: 2026-09-11 Categories: cs.LG, cs.AI, cs.SD, eess.AS arXiv: https://arxiv.org/abs/2609.12563v1 Neural audio codecs discretize speech into token sequences, but the resulting token spaces differ in vocabulary and codebook structure, preventing direct communication across models. This limitation affects applications such as conversational voice agents and speech to speech translation systems where multiple speech models must interact. As a result, transferring information between speech systems typically requires decoding to waveform audio and re-encoding with a second tokenizer, increasing latency and introducing potential information loss. To address these limitations, we present TokenMapper, a direction aware framework for direct token to token translation between heterogeneous speech tokenizers in the discrete domain. TokenMapper supports structurally mismatched token spaces, including mappings between single codebook and multi codebook representations, under a shared effective token rate. Experiments on GLM-4-Voice, MiMi and DualCodec show consistent cross model performance. Specifically, translation WER approaches native reconstructions within 2.5-6.8% absolute WER, human MOS for TokenMapper outputs ranges from 2.29 to 4.39, following the same direction level trends as UTMOS and end to end latency is reduced by 4.8-94.5% relative to waveform bridging, reaching up to 972 ms per utterance. These results provide a practical step toward cross model speech token interoperability without intermediate waveform reconstruction. --- ### KAD-Net: Kinematics-Aware Decoupled Learning for Robust 3D Hand Pose Estimation from a Single Depth Image Authors: Jun Lu, Zhenming Chen, Lin Chen, Kanlun Tan, Xiaoling Li, Qiao Liu Published: 2026-09-11 Categories: cs.CV, cs.HC arXiv: https://arxiv.org/abs/2609.12559v1 Due to the complexity of hand kinematics and self-occlusion, existing 3D hand pose estimation methods based on single depth images struggle to comprehensively model the topological dependencies among hand joints. Furthermore, traditional hierarchical multitask architectures enforce a shared feature space for both 2D joint localization and depth estimation, which can induce mutual interference. To address these challenges, we propose a Kinematics-Aware Decoupled Learning Network (KAD-Net) for robust 3D hand pose estimation. Specifically, we first design a Finger Topology Constraint (FTC) module to enhance the representation of distal joints. This module utilizes three consecutive finger joints to construct a local kinematic representation to impose topological constraints, which supplements the kinematic features of the distal joints. The FTC module leverages the structural context from visible joints to assist in locating occluded distal joints, thereby improving robustness to occlusion. Additionally, we propose a task-decoupled hierarchical multitask framework. This framework separates 2D joint localization from depth estimation and incorporates a dedicated multitask learning strategy for depth regression, effectively isolating the UV and depth features to mitigate mutual interference and negative transfer. Extensive experiments demonstrate that KAD-Net outperforms existing methods on several benchmark datasets (ICVL, NYU, and MSRA), achieving state-of-the-art accuracy in 3D hand pose estimation. Potential applications of KAD-Net include human-computer interaction, virtual reality and gesture-based control systems. --- ### DRS-VPT: Directly Relocalizing in a Scan with Vision Point Transformers Authors: Lanke Frank Tarimo Fu, Maurice Fallon Published: 2026-09-11 Categories: cs.CV, cs.RO arXiv: https://arxiv.org/abs/2609.12557v1 We present DRS-VPT, a feed-forward transformer architecture for foundational image-to-scan registration. Given query images and a reference 3D point cloud, the model predicts the scan pose and point map alongside the poses and point maps of each camera, all expressed in the first camera's frame. It additionally predicts a coarse-to- fine pyramid of per-point and per-pixel features for direct reprojective alignment of the scan to the first image. This formulation unifies downstream tasks such as camera-LiDAR calibration in autonomous driving and indoor camera-to-map relocalization. A single DRS-VPT model achieves state-of-the-art performance for image-to-LiDAR registration in autonomous driving, competitive indoor relocalization without training map-specific weights, and strong zero-shot transfer to unseen environments. We also show qualitatively that the model learns complex scan-to-image projection properties such as occlusion of back-facing points. --- ### RelateAnything: Real-Time Open-Vocabulary Relation Prediction From Any Inputs Authors: Maëlic Neau Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12552v1 Open-vocabulary detection accepts any class list at inference, and promptable segmentation returns regions without class names: the taxonomy has left the model and become an input. Relation prediction has not. Scene-graph models are still trained and evaluated on the 50 or 56 predicates of one annotation style, their relation head conditioned on object labels and so tied to one detector. Three obstacles explain this, none primarily modelling: no relation corpus is both free-text and verified, a label-conditioned architecture cannot accept a vocabulary it was not trained on, and the standard metric rewards agreement with the training corpus, so a larger vocabulary scores as a regression. We present RelateAnything, a 53M-parameter model taking an image and regions from any source and returning scored relations over a predicate vocabulary supplied at inference as strings. Object labels are never an input, so the region source can change without retraining, and the vocabulary is a bank of text embeddings, not a learned classifier. It runs at 20 ms/frame. Training over 19,103 predicates requires positive-unlabeled supervision and a text encoder that separates antonyms, which contrastive encoders embed at cosine 0.95. To supply the supervision we build RA-4M, 474k images and 4.3M relations over 10,102 free-text predicates, generated against numbered box markers and geometrically verified. To measure it we build OV-SGG-Bench, six axes scored across datasets that the priors standard recall rewards cannot satisfy. On three cross-dataset benchmarks and a fourth zero-shot, RelateAnything has 2.3-3.5x the mean recall of the strongest open-vocabulary method of comparable scale, margins that survive a real detector, and leads a 3B-VLM scene-graph model on both metrics at under 2% of its parameters. In-domain measurement overstates transfer gains ~5x. Model, corpus and benchmark are public. --- ### RoofLang: Enabling AI-Driven Architecting of LLM Inference Systems Authors: Ziyue Yang, Yuting Jiang, Lei Qu, Peng Cheng Published: 2026-09-11 Categories: cs.DC, cs.AI arXiv: https://arxiv.org/abs/2609.12551v2 AI is beginning to make substantive contributions to LLM inference optimization. Existing AI optimizations are predominantly profiling-based. Profiling-bound feedback confines the search to the capabilities and performance of an existing software stack, preventing a fundamentally better architecture of LLM inference systems from being identified. To enable the AI-driven LLM inference system architecting loop, we argue that a general workload representation, a verifiable mutation space, and an implementation-independent evaluator are required. We present the RoofLang domain-specific language (DSL) that provides these features. In our evaluation, RoofLang reveals that DeepSeek V4-series models could achieve 3.5-39.5$\times$ higher peak decode throughput than other representative models. This gap is disproportionate to their total parameter counts and arises largely from compact KV-cache designs that support larger batches and reduce memory traffic. A persistent optimizer agent further discovered several new architectures that improved both throughput and interactivity of DeepSeek V4 Pro on NVIDIA B300 by 6.23-50.1%. --- ### Quality-Constrained Routing over a Fixed Pool of Quantized Mixture-of-Experts Instances Authors: Zhenghong Huang, Hongfan Wu, Jiheng Zhang Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.12550v1 Quantized Mixture-of-Experts (MoE) services can hold several pre-materialized instances of one base model, but quantization damage varies sharply across requests and bitwidths. Because instance materialization and replica counts consume memory and require slow reconfiguration, we treat them as upstream provisioning decisions and study routing within a fixed resident pool. Within this fixed-pool boundary, we route each request to maximize modeled throughput under a class-level expected quality-degradation budget and measured instance capacities. To predict this request-specific risk, we introduce FWP (Fragility-Weighted Perplexity), computed from prompt tokens on a reference-instance prefill and calibrated to candidate-instance degradation. Underlying FWP is an exact two-expert affinity--fragility decomposition and a conditional multi-layer top-$k$ expansion whose bias, interaction, route-change, separability, and higher-order terms remain explicit. Using these calibrated risks, a window-level linear program yields a signed reduced-reward score that is KKT-consistent with the LP optimum under optimal prices and primal-feasible tie allocation. On 88 extended Qwen prompts, complete W2, W3, and W4 instances quantizing all 6,144 expert blocks incur mean $Δ$NLL of $0.9437$, $0.1832$, and $0.0513$. Under the same population and $τ=0.1513$, FWP allocation reaches a $1.284\times$ offline model-based multiplier versus $1.253\times$ for request-agnostic mixing and $1.000\times$ for static W4, an incremental $2.5\%$ relative FWP gain. --- ### Meddies-PII: A Multilingual Framework for Personally Identifiable Information Extraction in Clinical De-identification Authors: Linh Uyen Le, Christian Hoang, Huy Hoang Ha Published: 2026-09-11 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2609.12544v1 Clinical de-identification relies on accurately identifying personally identifiable information (PII). However, manually annotated datasets are costly to construct, while existing synthetic alternatives often provide limited details about their generation process or rely on relatively simple synthesis strategies. We introduce Meddies-PII-Dataset, a corpus of one million synthetic clinical documents spanning seventeen languages and nine PII labels. The documents are generated using attribute-conditioned prompts and validated through thirteen deterministic gates that enforce structural and annotation consistency. To evaluate the dataset's utility, we train Meddies-PII-Model, a BIOES token classifier, and compare it with existing PII extraction systems using exact-match entity-level F1. Meddies-PII-Model achieves the highest performance among the evaluated systems on all reported benchmarks, with a mean F1 of 0.827 across fifteen external benchmarks, compared with 0.658 for the strongest baseline. Upon acceptance, we will publicly release the dataset, benchmark suite, model, generation framework, and evaluation code to support research on multilingual clinical de-identification. --- ### Agent as Policy for Robotic Manipulation Authors: Mengzhao Jia, Yang Lin, Xixin Zhang, Zhihan Zhang, Xiaobai Liu, Meng Jiang Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.12541v3 We demonstrate that a general-purpose agent can directly drive a physical robot throughout task execution without any task-specific or environment-specific training. We introduce Agent as Policy (AGP), which places task planning and execution under the agent's control. Given a task and a robot interface, the agent interprets visual evidence, writes executable programs, issues motion commands, and revises its actions in response to physical outcomes. This brings the agent's reasoning and programming capabilities into continuous interaction with the physical world. We study AGP across multiple real-world manipulation tasks spanning precision manipulation, dynamic motions, and deformable objects. These include assembly from human videos, block construction from goal images, dice flipping, targeted throwing, and bimanual towel folding. AGP achieves success rates of at least 80% in seven of eight task configurations and significantly outperforms previous agentic robot systems. We further study efficiency through task experience accumulation and find that reusing saved procedures and programs shortens execution time across repeated trials. These findings support a path for general-purpose agents to act as robot policies, extending their autonomy to physical manipulation through runtime reasoning, programming, and interaction. --- ### The House with a Million Windows: Interactive Fiction for Narrative Restorying Authors: Cody Kommers, Sarah G Immel, Drew Hemment, Mina Lee Published: 2026-09-11 Categories: cs.CL, cs.HC arXiv: https://arxiv.org/abs/2609.12537v1 AI-assisted writing can flatten meaning in human storytelling, enabling the production of homogeneous outputs without the intentional effort and sense-making writing entails. To address this challenge, we present The House with a Million Windows (HWAMW), an LLM-based interactive fiction system designed to help users explore both the breadth and depth of potential meanings within their personal stories -- drawing on a psychological paradigm called the restorying intervention. In HWAMW, users play through a text-based narrative in which they tell a story, then encounter a set of LLM-generated "windows" reframing it according to different literary styles. Empirical evidence shows that HWAMW increases users' sense of narrative identity, while an expert review explores how this effect is achieved. Our findings suggest that HWAMW facilitates restorying and offers a valuable paradigm for AI-assisted writing, wherein LLMs do not tell our stories but rather help us see greater potential in the stories we tell. --- ### Earth-Agent-Pro: Towards Real-World Full-Chain Earth Observation with Agents Authors: Zhutao Lv, Chenhao Dang, Yi Feng, Yanpei Gong, Xiaolei Wang, Junyan Ye, Conghui He, Weijia Li Published: 2026-09-11 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.12533v1 Real-world Earth observation (EO) agents must translate high-level scientific questions into executable workflows to acquire observations, prepare data, perform domain computations, and derive conclusions from runtime evidence. Existing EO agents typically start from supplied observations, while benchmarks typically provide prepared inputs or candidate answers, leaving full-chain open-world EO execution largely untested. We present Earth-Agent-Pro, an execution-adaptive Plan-and-Execute framework using expert-authored skills to constrain planning and runtime tool use. Workflow-centered structured memory records planned steps, accepted evidence, and their dependencies, enabling repair of only the affected workflow suffix when runtime evidence invalidates a step. Separate large language model adapters use sequence-level supervised fine-tuning for planner workflow composition and node-level group relative policy optimization with locally verifiable rewards for executor tool-argument grounding. Earth-Bench-Pro instantiates 248 expert-curated task cores as 744 questions under three matched regimes. Its 248 Open-World Execution questions span RGB imagery, spectral observations, and remote sensing products, pairing high-level requests with runtime data requirements, executable trajectories, and open-ended answers grounded in execution evidence. With a shared GPT-5 backbone, Earth-Agent-Pro achieves 66.13% LLM-as-Judge accuracy, exceeding ReAct by 20.95 points in this metric and 24.44 points in Tools-In-Order. Joint adapter tuning raises Qwen3.5-9B LLM-as-Judge accuracy from 38.31% to 50.00%, an 11.69-point gain over the untuned configuration. Planning-only evaluation and execution with the reference workflow show that the adapters improve workflow composition and argument grounding, respectively. Code and datasets will be released soon. --- ### $\text{GSF-}χ$: Global Stereochemical Fields for Chiral Graph Transformers Authors: Jiaqing Xie, Yuxin Wang, Xipeng Qiu Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.12532v1 Enantiomers share atoms, bonds, and pairwise distances yet can behave differently in chiral environments, so molecular encoders must respect atom relabelings and proper rotations without becoming blind to reflection. We introduce GSF-$χ$, a graph transformer in which stereogenic units modulate all pairwise interactions rather than single out one atom as special. Each central or axial stereogenic unit creates a reflection-even phase field over all atoms, a handedness pseudoscalar $χ$ sets the direction of a relative rotation on latent query--key blocks, giving a \textbf{Chiral-RoPE} that reflection inverts rather than leaves fixed. A $C_2$ projection separates mirror-even ECD peak counts and positions from mirror-odd peak signs. We prove the operator's even--odd decomposition and its annotation-inversion, permutation, and unit-order identities under explicit canonical-role conditions; property tests and a coordinate-reflection audit verify the laws end to end. GSF-$χ$ leads every central-ECD output and improves axial Rotation and Symbol by $12.6\%$ and $7.9\%$ over the strongest baseline. Equal-budget controls attribute the Rotation advantage to global signed support rather than parameter or edge count; the $C_2$ projection yields exact enantiomer-pair consistency at a small raw-accuracy cost under complete supervision and becomes predictive when mirror supervision is scarce. --- ### Temporal Recurrence Favors Fewer Layers Authors: Ivan Anokhin, Johan Obando-Ceron, Irina Rish, Sebastian Risi Published: 2026-09-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2609.12531v1 In streaming tasks, recurrent models can carry latent computation across time, allowing each update to build on representations produced earlier. This raises a basic question: once temporal recurrence provides sequential computation across steps, how much depth is still needed within each step? Prior work has shown that recurrence can make shallow models competitive. We instead study this question as a compute-allocation problem, varying within-step depth, expert width, and the number of parallel experts per layer across several compute budgets. For each budget, we compare the best observed recurrent and non-recurrent allocations and the performance they achieve under approximately matched per-step computation. Across Sokoban and autoregressive FineWeb language modeling, we find that temporal recurrence shifts the best observed compute allocation toward substantially fewer layers, with comparable or better performance. --- ### Optimizing Geoengineering Interventions Using Differentiable Climate Models Authors: Pulkit Dubey, Dorian S. Abbot, Ashesh Chattopadhyay Published: 2026-09-11 Categories: physics.ao-ph, cs.AI, math.OC, nlin.CD arXiv: https://arxiv.org/abs/2609.12528v1 The deployment of a geoengineering program to cool Earth's climate may be imminent. It is crucial that tools be developed to ensure that such a program would achieve its objectives while minimizing disruption. Here we exploit recently developed differentiable atmospheric models to demonstrate a novel geoengineering control strategy. In the differentiable primitive-equation atmospheric model JAX-GCM we impose a uniform $+4$\,K ocean warming and ask what pattern of sea-surface temperature cooling -- in five ocean-masked zonal bands of prescribed SST forcings whose amplitudes are free -- returns land near-surface air temperature closest to the model's own unwarmed climatology. This idealized set-up represents a cooling pattern that could be delivered physically either by marine cloud brightening or stratospheric aerosol injection. Gradients through chaotic dynamics decorrelate from the true sensitivity beyond the Lyapunov horizon, so we optimize greedily over segments of 8 to 14 days, following receding-horizon control. The learned strategy removes $92.3 \pm 0.4\%$ of the realized land warming across a ten-member ensemble of two-year rollouts, and a three-year run sustains it. If we use the spatial pattern of land temperature as the optimization objective, the distributions of precipitation, evaporation, and specific humidity over land are restored as well, even though they are not included in the objective function. The learned strategy from JAX-GCM replayed in the AI emulators LUCIE and NeuralGCM without re-optimization is successful, suggesting robustness. These promising results demonstrate a strategy for designing optimal climate interventions that can be applied broadly for geoengineering scenarios under consideration. --- ### Learning-Augmented Optimization for Strategic Two-Echelon Spare Parts Network Design Authors: Donato Maragno, Marco Caserta, Alberto Sinigaglia, Komlanvi Ametana, David Corredor Montenegro, Luca D'Angelo Published: 2026-09-11 Categories: math.OC, cs.LG arXiv: https://arxiv.org/abs/2609.12524v1 We study the strategic design of a two-echelon spare-parts inventory network where evaluating each candidate topology requires an expensive inventory optimization model. The design partitions hundreds of sites into feasible clusters and selects a central replenishment site for each cluster to reduce costs while maintaining service levels. Because the optimizer favors candidates with high predicted savings, it can exploit optimistic surrogate errors. We develop a conservative framework combining a graph neural network ensemble, variable neighborhood search, and set-partitioning recombination. The surrogate is trained on exact cluster evaluations, while a lower quantile of ensemble-predicted savings guides the search to limit optimism. Clusters found during the search are recombined through set partitioning using surrogate-based objective coefficients. The resulting network is evaluated with the exact inventory model, and only this evaluation is used to report performance. In a case study of 246 fulfillment centers in Amazon's North American network, the framework improves combined savings by 30.5% over an optimization baseline based entirely on exact cluster evaluations, while maintaining approximately 99.8% service across six independent replications. Under equal computational budgets, graph-surrogate-guided search achieves higher mean exact savings than a tabular alternative under both scoring schemes. Conservative scoring improves mean savings for both surrogate classes and reduces the share of final-network clusters overestimated by the graph surrogate from 68% to 28%. Predictive and ranking accuracy deteriorate among search-generated candidates with high surrogate scores, indicating that random holdout performance can incompletely characterize surrogate quality during optimization. --- ### Aligned Radiometric RGB-Thermal Fusion for UAV Facade Anomaly Screening Authors: Yuan Yang, Shulei Li, Haobo Liang Published: 2026-09-11 Categories: cs.CV, eess.IV arXiv: https://arxiv.org/abs/2609.12521v1 Unmanned aerial vehicle facade inspection can combine red, green, and blue (RGB) imagery with thermal measurements to screen surface and subsurface anomalies. However, geometric discrepancies between the sensors and thermal image rendering can obscure spatial correspondence and weak temperature contrasts. This article presents a sensor-level pipeline comprising per-sensor correction, RGB-to-thermal registration, common-support cropping, and signed local contrast encoding of 16-bit radiometric measurements. The encoding preserves the distinction between locally hotter and colder regions and supplies the fourth input channel of a compact single-stream detector. We introduce M3T, a dataset of 674 paired RGB and radiometric thermal samples from five facade-inspection projects covering eight component and anomaly categories. The median residual registration error is 3.384 pixels, and a controlled-displacement analysis characterizes how the local contrast response changes under controlled displacement. Project-grouped four-fold evaluation yields mean average precision of 0.168 over intersection-over-union thresholds from 0.5 to 0.95, using 28.50 billion floating-point operations per image. A separate single-split ablation shows improved delamination detection over RGB-only and alternative thermal inputs, although aggregate accuracy does not improve over RGB alone. Evaluation on RGBT-Tiny shows mixed performance with rendered thermal imagery. These results characterize the category-specific benefits and limitations of aligned radiometric contrast for compact facade screening. --- ### One Skill Does Not Fit All: Automatic Discovery and Taxonomy-Guided Routing of Frame-Selection Skills for Long-Video Question Answering Authors: Jian Hu, Zixu Cheng, Da Li, Wei Li, Ziquan Liu, Shaogang Gong Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12517v1 Long-Video Question Answering (LVQA) requires locating decisive evidence in hour-scale videos under a limited frame budget. Most training-free methods apply the same frame-selection strategy to all questions, despite substantial variation in the evidence required by different question types. Our analysis shows that the relative effectiveness of frame-selection strategies varies across semantic categories and benchmarks, motivating adaptive evidence acquisition. In this paper, we introduce AutoSkill, a source-supervised framework for automatically discovering and routing executable frame-selection skills. Starting from a small labelled source pool, LLM agents iteratively propose, implement, evaluate, and refine candidate skills. For a target benchmark, AutoSkill uses only unlabelled question and option text to induce a shared semantic taxonomy, rewrite labelled source examples into the target style, and estimate a category-to-skill mapping. Neither target videos nor target answers are used in this process. At inference time, each question is assigned one skill, which selects the frames used in a single inference of the frozen video MLLM. Across five long-video benchmark splits, AutoSkill improves Qwen2.5-VL-7B and Qwen3.5-4B by 2.4% and 1.2%, respectively, demonstrating the effectiveness of our AutoSkill. --- ### A Splitting Method for SDE Terminal-Law Estimation Authors: Rushil Gupta, Sandeep Juneja Published: 2026-09-11 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2609.12513v1 In many settings involving stochastic differential equations, including in diffusion based generative AI, our aim is to accurately generate samples from a terminal distribution. Typically, this is done by generating i.i.d. samples of diffusion paths. Given a fixed simulation budget, a reasonable way to gain efficiency may be to instead generate a tree of paths through appropriately split partial paths. This suggests improved performance, but one worries about the injected dependence. In this paper, we study this issue comprehensively. With Kolmogorov-Smirnov distance as a measure of accuracy, we identify the limiting errors of the associated empirical distributions as the simulation budget increases to infinity. We characterize a splitting strategy motivated by a corresponding asymptotic optimization problem. The theoretical results bring out the elegant underlying structure in the problem. Practical implementation involves two phases, an initial estimation phase and a final inference phase. Overall, we observe a 10-25% improvement in mean error over i.i.d. samples in many settings. In an exploratory CIFAR-10 study, our method reduces the maximum mean discrepancy by 8-13%. --- ### Computing at Sea: Floating and Offshore Data Centres as a Pathway to Sustainable AI Infrastructure Authors: Cheng Siong Chin, Jianhua Zhang, M. Venkateshkumar Published: 2026-09-11 Categories: eess.SY, cs.AI arXiv: https://arxiv.org/abs/2609.12511v1 The rapid expansion of artificial intelligence is transforming data centres into one of the world's fastest-growing sources of electricity demand. As AI systems scale in size and capability, the physical infrastructure supporting computation is approaching critical limits in energy availability, cooling capacity, land use, freshwater consumption, and carbon management. Conventional land-based data centres are increasingly constrained by urban land competition, grid congestion, environmental pressures, and lengthy permitting processes, raising fundamental questions about where future computing infrastructure can sustainably exist. This article examines floating and offshore data centres as an emerging alternative model for digital infrastructure. By relocating computation to marine environments, offshore systems can exploit the ocean's natural cooling capacity, reduce freshwater dependence, and enable direct integration with offshore renewable energy resources such as wind, wave, and tidal power. Early deployments have demonstrated the potential for significantly improved energy efficiency and operational reliability compared with conventional facilities, while also opening new possibilities for distributed and resilient computing architectures. The article explores how offshore computing may reshape the future relationship between electrification, renewable energy, and large-scale AI infrastructure. It analyses the opportunities and trade-offs associated with marine deployment, including environmental impacts, engineering design challenges, economic feasibility, and regulatory governance. Rather than treating offshore data centres as experimental novelties, the article presents them as part of a broader systems-level transition in how society may power, cool, and sustain the next generation of computational growth. --- ### Signed Sensitivity of Expected Hitting Time to Mutation Rate in the (1+1) EA: Per-State Sign Theorems and Verifiable Certificates for Non-Lumpable Families Authors: RenKai Wang Published: 2026-09-11 Categories: cs.NE arXiv: https://arxiv.org/abs/2609.12510v1 For the (1+1) evolutionary algorithm with standard bit mutation, we study the sensitivity of the expected hitting time $H_p=\mathbb{E}_x T$ to the mutation rate. We first point out an easily overlooked formalization pitfall: the improvement event is not monotone in the mutation mask, so the unsigned (total-influence) form of the Margulis-Russo formula does not apply; the correct object is the signed endpoint difference. Second, we give an exact three-dimensional separation: two fitness functions share the entire one-step success-rate curve, yet their expected hitting times are two different exact rational numbers; hence one-step success-rate quantities do not determine the expected hitting time. Building on the runtime derivative $H'_p=(I-Q_p)^{-1}Q'_p H_p$, we construct computable double-residual sign certificates, prove a per-initial-state sign theorem on OneMax (for every non-optimal initial state, $\partial_c H<0$ on $0<c<1$, where $p=c/n$; at $c=1$ only the distance-one state is stationary), and extend the framework to non-lumpable positive linear families: an explicit non-lumpability witness, a block-interval double-residual certificate that covers all states without enumerating them, a uniform sign bound $\partial_c\mathbb{E}T\le -9n/16$ over the whole interval $c\in(1/4,1/2)$ for an explicit family at all even scales $n\ge 8$, and a heterogeneous instance certificate $H'_x\le -1/6$ on 57 of 63 states across $c=1$. All finite verifications use exact rational arithmetic. A bounded systematic literature search did not uncover this exact combination, although the underlying tools are well established; we therefore make no novelty claim beyond the stated combination. --- ### ChitraMiti: Benchmarking Visual Grounding and Modality Reliance in Bengali Geometric Reasoning Authors: Khan Raiyan Ibne Reza, Sanjana Aktar Maria, Sumaiya Tabassum Nimi, Md Adnan Arefeen Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12509v1 Evaluation of vision-language models (VLMs) for multimodal mathematical reasoning remains limited for low-resource languages and for geometry problems that require reading a diagram and a question together. We introduce ChitraMiti-12.8k, a synthetic benchmark of 12,874 Bengali planar geometry problems paired with structured 15-attribute descriptions, and NCTB-500, a complementary set of 500 diagrams manually extracted from Bengali school textbooks. Using a three-phase protocol that separates diagram-only, diagram-plus-description, and description-only inputs, we show across five open-weight and closed-source VLMs that description-only performance is statistically indistinguishable from diagram-plus-description performance, establishing structured descriptions as a sufficient textual proxy for controlled evaluation. Despite this, models remain poor at cross-modal verification, frequently misled by a swapped spatial relation even when they answer the unmodified item correctly. We further evaluate supervised adaptation on ChitraMiti-12.8k, finding that fine-tuning improves performance on both ChitraMiti-1k and NCTB-500, although a substantial gap to the strongest zero-shot model remains. Together, ChitraMiti-12.8k, NCTB-500, and our evaluation protocol offer a standardized way to study Bengali multimodal geometry reasoning and, more broadly, whether VLMs actually check their text against what they see. Our dataset and code are publicly available on Hugging Face at https://huggingface.co/datasets/RaiyanKhaan/ChitraMiti. --- ### LettuceVisSim: A Simulator That Generates Lettuce Image Time-series for Vision-Based Reinforcement Learning Authors: Ziye Zhu, Bert van 't Ooster, Congcong Sun, Eldert van Henten, Sjoerd Boersma Published: 2026-09-11 Categories: cs.CV, cs.AI, eess.SY, math.OC arXiv: https://arxiv.org/abs/2609.12505v1 Vision-based reinforcement learning holds strong potential for decision-making in controlled environment agriculture (CEA). However, its development is hindered by the scarcity of labelled crop images. To address this gap, LettuceVisSim, a lettuce growth simulator that generates labelled time series of crop images, was developed and validated. The simulator contains a process-based model (PBM) for shoot dry weight dynamics, a canopy layout algorithm for deriving canopy layout representations from shoot dry weight, and a Unity rendering engine for image generation. Five findings support the simulator. First, the PBM reproduced shoot dry weight under dynamic plant-density management with $\mathrm{R}^{2}=0.84$. Second, a piecewise cubic regression mapped shoot dry weight to potential projected area with $\mathrm{R}^{2}=0.94$. Third, the canopy layout representation was validated using 12 experimental datasets each having different dynamic environmental and spacing conditions. It reproduced the ground coverage ratio dynamics observed in measured images, achieving $\mathrm{R}^{2}=0.84$ when driven by measured shoot dry weight and $\mathrm{R}^{2}=0.40$ (0.76 excluding one outlier) when driven by PBM-simulated values. Fourth, the Unity rendering engine converted canopy layout representations into RGB and segmentation images at less than 10~ms. Fifth, a demonstration showed that a lighting-control policy can be learned and applied by observing only crop images that were generated with LettuceVisSim, providing a proof of concept of vision-based reinforcement learning in CEA using LettuceVisSim. --- ### When2Talk: When Should a Proactive In-Car Agent Talk? Authors: Kaiser Hamid, Peihang Li, Nade Liang Published: 2026-09-11 Categories: cs.HC, cs.CV arXiv: https://arxiv.org/abs/2609.12503v1 Proactive in-cabin agents can help passengers understand automated-vehicle (AV) behavior, but communicating every ride event may introduce unnecessary interruptions. We investigated how communication should adapt to event priority and passenger activity. In a mixed-methods within-subject study, 41 participants rode as passenger in a VR simulated fully-automated vehicle. We compared an event-triggered (ET) policy that communicated immediately at every event with a context-sensitive (CS) policy that selected \textit{Immediate}, \textit{Delayed}, or \textit{Silent} communications. CS increased communication appropriateness and substantially reduced perceived interruption. Perceived trust did not differ between policies, although baselines dispositional trust differentiated communication preferences. Findings highlight event consequence, passenger activity, continuing information value, and confirmation need as key considerations for selective in-cabin communication. --- ### RoES: Rotational Equivariant Selective-frequency Fusion for Multimodal Images Authors: Jiabao Wang, Wenjian Liu, Yaoming Cai, Gengyu Zhang, Boyan Zhao, Zijia Zhang, Yao Ding, Xiaobo Liu Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12497v1 Infrared-visible image fusion facilitates robust multimodal perception by integrating complementary textural nuances from visible sensors with thermal signatures from infrared systems. Due to the task's inherently ill-posed nature, existing methods heavily rely on structural priors but typically enforce rotation equivariance uniformly across all features. Such a holistic approach overlooks a critical distinction where low-frequency shared structures strictly adhere to equivariant constraints while high-frequency modality-specific details require greater flexibility to preserve unique information. To bridge this gap, we propose RoES, a Rotational Equivariant Selective-frequency fusion network. Instead of employing static decomposition, we introduce a trainable rotation-enhanced updater/predictor module to dynamically decouple low- and high-frequency components. The resulting representations are then processed through a dual-branch fusion module tailored for spectral consistency. Specifically, a rotation-equivariant Mamba is employed to capture long-range structural dependencies in the low-frequency domain, while a polar spectral attention-based Dual-Fourier block refines high-frequency details under explicit low-frequency guidance. Extensive experiments demonstrate that RoES consistently achieves state-of-the-art performance in both fusion quality and downstream object detection, establishing a robust solution for multimodal fusion by reconciling frequency-selective features with equivariant constraints. The source code is available at https://github.com/BryceLosky/RoES-Fusion. --- ### Information Specialization and Constrained Synthesis in Multi-Agent LLM Forecasting: A Prospective Live-Study of the 2026 FIFA World Cup Authors: Julian Varghese, Lucas Bickmann, Sarah Sandmann Published: 2026-09-11 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2609.12495v1 Large language models are being organized into multi-agent systems with specialized roles, but whether such specialization produces distinct forecasts and whether subsequent synthesis improves utility remains unclear. In this study, we carried out a live, prospective evaluation over the final 56 matches of the information-dense 2026 FIFA World Cup, keeping a frontier foundation model constant while assigning two primary forecasting agents contrasting specialist roles: a quantitative specialist focusing on structured performance statistics and a news specialist focusing on current injuries, tactics and information from press conferences. Their forecasts were then reviewed by a separate critic before being combined by a meta-agent, resulting in a sequential four-agent model. Forecasts from the betting market served as an external benchmark. The news specialist obtained the highest mean probability-weighted Top-3 utility and matched the betting market in Top-3 exact-score hits. Nevertheless, the two specialist forecasters agreed on at least two of the three scorelines in 50 out of 56 matches, and the meta-agent never generated more than one scoreline outside the specialists' forecast set. These findings show that rapidly changing, unstructured information can provide a valuable forecasting signal alongside structured statistics, whereas adding critic and meta-agent stages does not necessarily create complementary information or improve on the strongest specialist. --- ### PhysioAI: Clinical Knowledge-Guided Semantic Supervision for Skeleton-Based Physiotherapy Action Recognition Authors: Jie Cao, Euijoon Ahn, Anwar Hassan, Jinman Kim Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12491v1 Skeleton-based action recognition can support automated tracking of physiotherapy exercises, particularly in remote rehabilitation settings where continuous in-person supervision is impractical. However, most existing methods are developed for large-scale daily-action benchmarks rather than rehabilitation scenarios. Public rehabilitation exercise datasets are typically small, with only subtle kinematic differences between exercise classes. For participants with motor impairments, exercise execution may also deviate from standard movement patterns in amplitude, speed, and coordination, increasing intra-class variability and making reliable recognition more difficult for skeleton-based models. We propose PhysioAI, a clinical knowledge-guided semantic supervision framework that injects structured physiotherapy knowledge into skeleton representation learning. PhysioAI combines graph-based spatiotemporal modelling of human movement with training-time semantic anchors derived from a structured Clinical Knowledge Dictionary (CKD). The CKD descriptions are encoded using a frozen Contrastive Language-Image Pre-training (CLIP) model and projected into an anchor space, where they provide class-specific semantic targets for skeleton representation learning. The resulting CKD-derived anchors are used only during skeleton-model training; inference requires only skeleton inputs. Under subject-disjoint evaluation, PhysioAI achieves $99.03\pm1.34\%$ on KiMoRe Overall, $94.64\pm7.36\%$ on the Hard-67 stress test, and $87.44\pm7.69\%$ on UI-PRMD Overall. These results exceed the strongest comparator for each endpoint by $0.27$, $2.87$, and $1.33$ percentage points (pp), respectively. These findings demonstrate that structured clinical knowledge can serve as an effective source of training-time supervision for physiotherapy action recognition. --- ### Confidence-Gated Transductive Test Generation for Code Reranking Authors: Sungjae Lee, Youngsik Yoon, Seockbean Song, Siwei Wang, Wei Chen, Jungseul Ok Published: 2026-09-11 Categories: cs.AI, cs.CL, cs.SE arXiv: https://arxiv.org/abs/2609.12489v1 Test case synthesis is crucial for evaluating and ranking programs generated by large language models (LLMs). However, constructing high-quality test cases remains challenging because reliable expected outputs are often difficult to obtain. We propose Confidence-Gated Transductive Test Generation (CoTT), which first uses an efficient inductive procedure and invokes transductive generation only when inductive confidence is low. This adaptive design improves output reliability while allocating extra computation only when needed. On code reranking benchmarks, CoTT outperforms prior baselines across the reported metrics while reducing cost relative to applying transductive generation to every input. These results show that confidence-based allocation of test-time computation provides a favorable efficiency-effectiveness trade-off with a single efficient LLM. --- ### Adaptive Agent Design Authors: Raj Kiriti Velicheti, Subhonmesh Bose, Tamer Başar Published: 2026-09-11 Categories: eess.SY, cs.AI, cs.GT, math.OC arXiv: https://arxiv.org/abs/2609.12486v1 We consider an agent acting against a general non-Markovian environment. The agent maintains its agent states, but is free to choose a transition kernel across those states and optimize its state-feedback control policies. We study the bi-level agent design problem that optimizes the transition kernel and the policy it induces, given said kernel with offline data of observations and actions obtained via a behavioral policy. For general environments, we show that a soft $Q$-learning algorithm converges almost surely to the fixed point of a soft Bellman equation defined by the stationary averages that the behavioral policy and the chosen kernel induce, and we delineate what separates the resulting policy from an optimal one. In partially observed Markov decision problems, we analyze convergence properties of parametrized transition kernel design via zero-th order and Bayesian optimization techniques. --- ### When Does AI Augment Work? A Workflow-Level Framework for Human-Agent Collaboration Authors: CIVIC-AI Collaboration, :, Jiaying Wu, Caleb Ziems, Raymond Chan, Nancy F. Chen, Corlyss Chua, Gerard Chung, Jungpil Hahn, Wee Sun Lee, Zhengyuan Liu, Jamie Ng et al. Published: 2026-09-11 Categories: cs.AI, cs.CY, cs.HC arXiv: https://arxiv.org/abs/2609.12482v1 We aim to characterise the value of artificial intelligence in the workplace. Current studies largely measure this value in terms of the current automation capabilities and public adoption of AI. However, such metrics ignore the greater impacts of human--agent collaboration in transforming the nature of work. To account for this, we must expand the scope of our analysis beyond atomised tasks of today, and instead focus on how AI can augment entire workflows of the future. To ground this analysis, we establish a precise definition of AI augmentation comprising six conditions, spanning durable net value, meaningful human control, accountability and recovery, and long-term human development through learning, career pathways, and job purpose. We elaborate on these conditions and apply the framework in a case study of AI-mediated social surveys. We conclude by outlining how organisations, researchers, and government leaders can use this framework to make sense of the future of work. --- ### Zipbench: Low-Cost Framework for Compressing Comprehensive Benchmarks of Large Language Models Authors: Zhongzhan Huang, Junxin Li, Guoming Ling, Yupei Lin, Shanshan Zhong, Hefeng Wu Published: 2026-09-11 Categories: cs.CL arXiv: https://arxiv.org/abs/2609.12475v1 Comprehensive benchmark suites are essential for improving large language models (LLMs), but many widely used benchmarks are redundant, making evaluation unnecessarily expensive. Although recent benchmark compression methods (BCMs) can mitigate this cost, many strong BCMs rely on large collections of per-sample evaluation results from numerous LLMs to identify representative samples. Building such collections is also expensive unless they are already public, making these methods difficult to extend to newly released benchmarks. To address this challenge, we present ZipBench, a simple and low-cost BCM with theoretical error and rank-consistency guarantees. ZipBench evaluates only a small set of anchor LLMs, synthesizes pseudo evaluation results to broaden coverage, learns compact sample representations, and selects a small yet representative subset. Building on it, we create ZipBench Zoo, a collection of compact versions of 100+ benchmark proxies spanning text, multimodal, and agent tasks. These benchmark achieve mean absolute errors of 0.002--0.02 and average Spearman correlations of ~0.98 with the full benchmarks. Overall, ZipBench reduces the cost of both LLM evaluation and compact benchmark construction, lowering the barrier to broad LLM research for compute-constrained researchers. The code has been released in https://github.com/MilkThink-Lab/ZipBench. --- ### Partition-Invariant Tuning for 3D Scene Understanding Authors: Hongqiang Lin, Tianle Wang, Shuiwang Li, Dongxu Zhang, Yiding Sun, Zihao Guo, Dongfu Yin Published: 2026-09-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2609.12473v1 Scene-level point cloud understanding remains challenging due to diverse geometries and spatial layouts. While pre-trained 3D point cloud foundation models (PFMs) offer strong transferability, full fine-tuning (FFT) incurs substantial computational and storage costs. Parameter-efficient fine-tuning (PEFT) provides a promising alternative, but existing PEFT methods largely focus on object-level point clouds and overlook serialization-induced partition variations in large-scale scenes. To address this issue, we propose PointPiT, a partition-invariant tuning framework for scene-level point clouds. Specifically, a Scene-aware Structural Adapter (SSA) integrates local geometric patterns with global scene context to mitigate partition-induced representation shifts. Moreover, Gradient Subspace Optimization (GSO) selects informative and partition-stable update directions, suppressing partition-dependent variations during optimization. Extensive experiments across multiple scene-level benchmarks demonstrate that PointPiT achieves competitive or even superior performance to full fine-tuning with less than 1% of backbone's parameters, while achieving consistent state-of-the-art performance among representative PEFT methods. --- ### TripPattern: A Pattern-based Text Watermarking Method for Large Language Models Authors: Sangjun Moon, Dasom Choi, Jingun Kwon, Hidetaka Kamigaito, Taro Watanabe, Manabu Okumura Published: 2026-09-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2609.12472v1 Text watermarking techniques have gained significant attention for identifying machine-generated text and mitigating risks from large language models (LLMs). Existing methods typically divide an LLM's vocabulary into green and red tokens, but encouraging generation toward green tokens can reduce text quality and naturalness. To address this, we propose TripPattern, a watermarking framework that formulates text watermarking as a pattern-based matching task using three vocabulary partitions. TripPattern divides the vocabulary into one neutral group and two pattern groups. During generation, the model alternates token selection between the two pattern groups to embed detectable patterns, while neutral tokens are selected independently to improve flexibility and preserve naturalness. For detection, TripPattern uses pattern-based statistical tests that provide interpretable p-values by measuring how often adjacent tokens alternate between the pattern groups. Theoretical analysis and empirical evaluations on four multilingual datasets show that TripPattern maintains LLM generation quality while achieving robust watermark detectability. ---