# Research Papers — 2026-10 · bundle page 121 of 156 > Full-text markdown of papers #6001–6050 of 7799 published in 2026-10, newest first. AI-readable mirrors of the full 7799-paper 2026-10 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-10/page/121 [← Previous bundle](https://aipod.dev/papers/archive/2026-10/page/120.md) [Next bundle →](https://aipod.dev/papers/archive/2026-10/page/122.md) ### Hyperbolic Graph Representation Learning: Embed in One Metric, Optimize with Another Authors: Federico Larroca, Paola Bermolen, Marcelo Fiori, Bernardo Marenco Published: 2026-10-05 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2610.06745v1 Hierarchical graphs embed in hyperbolic space with lower distortion than in Euclidean space owing to its negative curvature. However, their gradient-based learning is hampered at large radii, where the Poincaré ball and the Lorentz hyperboloid models fail numerically. Polar coordinates avoid this problem, but the hyperbolic metric scales the angular step by the hyperbolic sine of the radius, freezing angular motion. We observe that this factor is a choice, silently fixed by existing implementations: the Euclidean tangent parametrization, for instance, uses the radius itself. We show that other choices are not only possible but preferable. They are endpoints of a one-parameter family of optimization preconditioners with curvatures from $-1$ to $0$, while the embedding remains at curvature $-1$. We show that since the Euclidean preconditioner rearranges a layout but refines it poorly, while an intermediate one refines far better once a layout is in place, combining them in two stages reduces the loss on real-world trees by 46-74% over the best single curvature. --- ### ufakzeka-karar: An Open Turkish Typed-Decision Model with Order-Invariant Option Scoring Authors: Sait Furkan Teke Published: 2026-10-05 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2610.06744v1 ufakzeka-karar is an open Turkish decision model with 182,494,466 parameters. Given a Turkish text and questions of a fixed answer type (a choice, a level on an ordered scale, or yes or no), it returns a temperature-scaled probability for every option and an expected error that serves as a "not sure" signal, without generating text and in one CPU forward pass for up to ten options. Built on the lab's ufakzeka-1-base, its head scores each option blind to the others at shared positions, so the answer does not depend on option order. A sequential head trained with shuffled options was about as accurate but changed 2.3 to 2.8 percent of its answers when only the option order changed; REINFORCE lost 10.2 points (0.102) of macro F1 to cross-entropy. On the open set of HakemBench v1.0 (4,275 questions, 7 tracks) the released model ranks 7th of 16 rows with a composite of 0.660 (95% interval 0.642 to 0.677). Temperature scaling lowers calibration error (smooth ECE) on the development set but raises it on held-out support questions, from 0.027 to 0.045 for the first scored run, which never trained on them; the released model later trained on them, so its 0.036 to 0.064 is not an unseen-question test. The released model is the last of three runs scored on HakemBench, and its numbers are not blind. The second run's new training data was aimed at the first run's errors on the full test set in guardrails, moderation and customer support, and the released run was trained after the second run's guardrail results on the full test set were read, under a protocol fixed in writing before any of its data, code or runs. All its numbers come after these readings; its guardrail, moderation and customer support numbers carry the flag "shaped by reading the test results". With every model scored on the other four tracks only, its composite is 0.678, 6th of 16. Weights and code are under Apache-2.0. --- ### Improving Diversity in LLM Short Story Generation Authors: Zahra Solati Dehkordi, Vasileios Lampos Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06729v2 Large language models (LLMs) can generate accurate responses, but these are void of diversity. We attempt to address this for the task of creative short story generation. Drawing on established writing conventions and known LLM limitations, we target variation in genre, tone, style, and named entities. To promote diversity across these dimensions, we introduce DivLM, an LLM post-training framework consisting of two phases. First, we perform continued pre-training on a creative writing corpus and restore instruction-following capabilities using weight residuals. We then apply reinforcement learning with a custom, composite reward function that jointly maximizes diversity across the targeted narrative dimensions while maintaining response quality. Our empirical results on two LLM families show that DivLM increases diversity metrics by more than 9% on average compared to alternative approaches, while preserving instruction following, overall response quality, and similarity to human outputs. --- ### Improving Diversity in LLM Short Story Generation Authors: Zahra Solati Dehkordi, Vasileios Lampos Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06729v1 Large language models (LLMs) can generate accurate responses, but these are void of diversity. We attempt to address this for the task of creative short story generation. Drawing on established writing conventions and known LLM limitations, we target variation in genre, tone, style, and named entities. To promote diversity across these dimensions, we introduce DivLM, an LLM post-training framework consisting of two phases. First, we perform continued pre-training on a creative writing corpus and restore instruction-following capabilities using weight residuals. We then apply reinforcement learning with a custom, composite reward function that jointly maximizes diversity across the targeted narrative dimensions while maintaining response quality. Our empirical results on two LLM families show that DivLM increases diversity metrics by more than 9% on average compared to alternative approaches, while preserving instruction following, overall response quality, and similarity to human outputs. --- ### Decoupling Time and Space: A Temporally Conditioned Refinement for EEG Source Imaging Authors: Marco Morik, Jesse Palarus, Carmen Vidaurre, Klaus-Robert Müller, Shinichi Nakajima Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06726v1 Electroencephalography (EEG) offers millisecond temporal resolution, but inferring underlying neural sources is a severely ill-posed spatial inverse problem. While deep learning has advanced spatial reconstruction, current architectures face a critical dilemma: frame-by-frame models discard vital temporal context, whereas full 4D spatiotemporal networks introduce an architectural trade-off between reconstruction accuracy and inference cost. We propose a novel two-stream framework that explicitly decouples global temporal representation learning from per-time-point spatial refinement. A Transformer-based Temporal Condition Encoder processes the entire EEG sequence via factorized spatiotemporal attention, retaining sensor-resolved features. A fixed inverse then maps these features into source-indexed conditioning for a per-timestep Source-Space Transformer or volumetric convolutional refiner. Extensive evaluations on realistic synthetic data demonstrate that this temporal prior dramatically improves spatial localization, outperforming classical and spatiotemporal baselines, particularly in high-noise and multi-source regimes. Training across diverse leadfields and explicit operator mismatches improves transfer to unseen head geometries and brings template-based reconstruction closer to subject-specific inversion. Furthermore, we apply the model trained only on synthetic EEG data to real-world EEG. A logistic regressor fit on source power differences in eyes-open, eyes-closed conditions successfully decodes age groups. --- ### BRANCH-MoE: Balance-Aware Tree Routing for Large Embedding Models Authors: Gang Fu, Adel Javanmard, MohammadHossein Bateni, Vahab Mirrokni Published: 2026-10-05 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2610.06725v1 Mixture-of-experts (MoE) layers increase model capacity without a proportional increase in per-example computation. However, conventional flat routers can yield imbalanced expert utilization and treat experts as an unstructured collection, whose indices carry no topological meaning. We introduce {\bf BRANCH-MoE}, a routing architecture that places \(E\) experts at the leaves of a binary decision tree of depth \(\log_2 E\). At each internal node the branching probability is centered on the arrival-weighted mean score of the traffic reaching that node. This mean is estimated using an exponential moving average, which promotes utilization of both child subtrees without an auxiliary load-balancing loss. We show that this moving-average estimate admits an explicit noise-lag trade-off. We prove that for linear node maps and log-concave arrival distributions, this mechanism prevents routing-mass collapse. We further establish that, under a frozen router, an expert's execution frequency controls its stochastic-gradient convergence rate, and that confident decisions near the root bound cross-device communication when experts are assigned to devices by tree prefix. We evaluate BRANCH-MoE against Switch softmax, DeepSeek-V3 dynamic-bias, Skywork logit-normalized, and deterministic hash routing on Criteo click-through-rate prediction, Forest Covertype, HIGGS, and YearPredictionMSD, using \(E=16\), top-\(4\) routing, and five random seeds. Our results show that hierarchical routing can preserve task quality and balanced utilization while inducing a topology that supports localized expert co-activation and reduced communication. --- ### Domain adaptation of Russian ModernBERT for long legal documents Authors: I. Litvak, D. Gvozdetsky, F. Lashkin, V. Kirova, S. Lagutin, V. Volf, T. Maksiyan, A. Kostin, R. Leva, I. Kiselev Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06715v1 We investigate whether continued pretraining on Russian legislative documents improves a Russian ModernBERT encoder on legal text. The adapted model, RuModernBERT-ruLaw, was trained on a corpus reported to contain 304,382 legislative documents and 194,425,905 corpus tokens. Corpus token counts are distinguished from positions produced by the model tokenizer. We compare the original and adapted encoders on a fixed external collection of 1,031 court-decision segments. Both models receive the same hidden positions in each of five masking realizations. At maximum input lengths of 512, 2,048, and 8,192 tokens, mean masked-token cross-entropy decreases by 0.10942, 0.07052, and 0.06604 natural-log units, respectively. The reported 95% intervals summarize sensitivity to masking on this fixed collection; they do not quantify uncertainty across document collections. A second evaluation addresses legal-entity extraction. The original and adapted models achieve entity-level F1 scores of 0.99852 and 0.99820. However, 99.95% of test spans have the same normalized surface form and class in the training split. This evaluation therefore provides limited evidence about transfer to previously unseen forms. The paper explains the masking objective, overlapping windows, averaging rules, and exact entity-boundary scoring using editable diagrams and clearly marked illustrative examples. The comparison supports lower masked-token prediction loss for the studied pair of models and collection. It does not isolate the contribution of distant context or establish practical legal utility. --- ### Out-of-control Hamiltonian Learning Authors: Weiyuan Gong, Muzhou Ma, Sitan Chen, Jordan Cotler, Hsin-Yuan Huang Published: 2026-10-05 Categories: quant-ph, cs.IT, cs.LG, physics.atom-ph arXiv: https://arxiv.org/abs/2610.06709v1 Learning the Hamiltonian of a many-body system from its dynamics is a central task in quantum science, yet the algorithms with the strongest provable guarantees assume some level of quantum control--fast, arbitrary single-qubit gates interleaved with time evolution, and measurements in arbitrary bases--that is beyond the capabilities of near-term analog quantum simulators. Motivated by analog atom- and ion-based platforms, we study Hamiltonian learning under minimal access models. Uniform state preparation and measurements: We first consider the setting where in every experiment, one can rotate each qubit to the same state, perform short-time evolution, and measure every qubit in the same basis. Surprisingly, we show that for generic 2-local Hamiltonians on any interaction graph, all of the parameters can be reconstructed from such experiments. Computational basis state preparation and measurements: We then consider a similarly constrained setting, but where state preparation and measurement are restricted to the computational basis. For nearest-neighbor Hamiltonians with only Pauli $X/Z$ interactions, a class which captures contemporary Rydberg atom platforms, we show that over 1D and 2D rectangular lattices, all of the parameters can be reconstructed from such experiments up to unavoidable gauges. Our protocols introduce new techniques for solving structured polynomial systems over an extensive number of parameters. Taken together, our results suggest that one can learn a great deal from the dynamics of quantum many-body systems even under the most stringent experimental constraints. --- ### SAFE-MR: Evidence Sufficiency Learning for Selective Multimodal Rumor Detection Authors: Shiwen Ni Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06708v1 Multimodal rumor detectors increasingly rely on retrieved evidence, yet relevant evidence is not necessarily sufficient for verification. Missing provenance, duplicated reports, and unresolved contradictions can produce confident predictions without adequate support. We introduce SAFE-MR, a framework that separates claim veracity from evidence sufficiency. The method decomposes image-text posts into verifiable claims, constructs a relation-aware claim-evidence graph, and aggregates evidence using provenance and contextual compatibility. Separate veracity and sufficiency heads support selective prediction, while evidence interventions encourage stability under irrelevant additions and sensitivity to evidence removal. On NewsCLIPpings, VERITE, and XFacta, SAFE-MR achieves macro-F1 scores of 91.2%, 75.8%, and 85.2%, respectively. Against the matched backbone with evidence, its macro-F1 gains are 2.2, 4.9, and 4.8 percentage points. On the diagnostic selection set, SAFE-MR reduces AURC from 0.105 for maximum-probability rejection to 0.075 and lowers error at 80% coverage from 13.8% to 8.5%. Evidence-perturbation and ablation results support the role of sufficiency learning and intervention training in improving selective verification. --- ### Extending Dynamic World Surface Water Mapping to Sentinel-1 with AlphaEarth Embeddings Authors: Rohit Mukherjee, Frederick Policelli, Beth Tellman, TC Chakraborty, Jonathan Giezendanner, Jonathan A. Sullivan, Ning Sun Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06704v1 Dynamic World (DW) maps land use and land cover globally at 10 m from Sentinel-2 (S2) imagery, but only for cloud-free observations, which limits where and when surface water can be mapped. We use the DW water class as weak supervision for a Sentinel-1 (S1) synthetic aperture radar (SAR) model so that DW-like water maps can be produced for every S1 acquisition. Google's AlphaEarth Foundations (AEF) annual embedding supplies spatial context, while S1 backscatter supplies the acquisition-time observation. On 53 globally distributed scenes with independent annotations of 3 m PlanetScope imagery acquired within 48 h of the S1 overpass, the S1-only model already reaches a pooled water intersection over union (IoU) of 0.77, comparable to 0.75 for the operational OPERA DSWx-S1 product, and adding AEF raises it to 0.85. The fused model improves on the S1-only model on 44 of 53 scenes and exceeds OPERA on 48, and on the independent S1S2-Water benchmark it reaches 0.94, compared with 0.87 for OPERA. Optical land-cover products can thus provide scalable training labels for SAR surface water mapping. --- ### Reading the Mood: Emotion-Guided Book-to-Music Recommendation via CGANs and LLMs Authors: Manousos Linardakis, Georgios Alexandridis Published: 2026-10-05 Categories: cs.IR, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2610.06703v1 Background music that matches the mood of a text has been shown to make readers feel more immersed and improve their reading experience, motivating recommender systems that pair books with mood-matched music. In this direction, we present Sentiment Aware Generative Adversarial Network for Cross Domain Recommendation (SAGA-CDR), a two-phase cross-domain recommendation framework that personalizes music suggestions and emotionally aligns them with the book being read. In the first phase, transformer-based sentiment embeddings are constructed from user reviews and mapped across domains via a Conditional Generative Adversarial Network, whose mask-conditioned generator handles missing sentiment components and injects stochasticity for richer preference transfer. A compact rating neural network then fuses sentiment-specific interaction scores with a collaborative filtering prior to predict music ratings. In the second phase, large language models classify each book into a valence-arousal emotional quadrant, and candidate tracks are filtered to match that quadrant. Experiments on both the English Amazon and Chinese Douban datasets show that SAGA-CDR achieves the best rating prediction accuracy on Amazon (RMSE 0.98) and the lowest RMSE on Douban (0.91), with ranking performance competitive with the strongest sentiment-aware baseline, even in cross-lingual settings. --- ### A Solvable Model of Adaptive Learning Rate Rescaling: Acceleration, Stability & Scaling Authors: Itay Lavie, Clarissa Lauditi, Cengiz Pehlevan Published: 2026-10-05 Categories: stat.ML, cond-mat.dis-nn, cs.LG arXiv: https://arxiv.org/abs/2610.06701v1 A recurring design principle in modern optimizers is to decouple update magnitude from the raw gradient norm, yet its consequences for learning-curve and resource scaling remain unclear. We isolate this mechanism by studying normalized SGD in a random-feature model with power-law teacher and data covariance. Fixed-norm updates induce an effective learning rate that grows as gradients shrink. We derive a dynamical mean-field theory (DMFT) describing the joint dependence of the loss on training time, model width and batch size. Normalization initially accelerates SGD, mapping the power-law exponent $r_{\rm SGD}<1$ to $2r_{\rm SGD}/(1-r_{\rm SGD})$, with exponential convergence at $r_{\rm SGD}=1$ and formal finite-time convergence for $r_{\rm SGD}>1$. At finite step size, however, the same feedback ultimately breaks the acceleration and leads to marginal stability. The late-time theory yields width-limited, edge-of-stochastic-stability (EoSS), and deterministic edge-of-stability (EoS) regimes. These phases determine when larger batches or wider models reduce serial training time at comparable compute. We quantify in which of these phases increased batch size or width can compensate the excess compute use per step by fewer optimization steps to target loss. Linearized ResNet experiments on CIFAR-5M support the predicted acceleration, breakdown, and resource-scaling trends. Together, these results connect normalization-induced acceleration, EoS effects, and width--batch allocation within a solvable theory. --- ### MedPrune: Topology-Efficient Multimodal Multi-Agent Communication Evolution for Medical VQA Tasks Authors: Jiuheng Wan, Runze Li, Chen Chen, Tingyuan Hu, Daiyang Yu, Yimin Jing, Taolin Zhang, Richang Hong Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06695v1 While medical multimodal large language models (Med-MLLMs) advance medical visual question answering (VQA), existing clinical workflow-inspired multi-agent frameworks suffer from interaction patterns and excessive computational overhead caused by redundant communication topologies. In this paper, we propose MedPrune, an efficient medical multimodal multi-agent collaboration framework that dynamically prunes both nodes and edges from the communication topology to enhance reasoning ability and token efficiency. Specifically, we first formulate the diagnostic process as a heterogeneous communication graph, where nodes represent specialist agents from various departments and edges capture intra- and inter-departmental interactions. Building on this graph, we introduce two sparsification mechanisms to enable adaptive collaborative evolution: (1) Heterogeneous Node Sparsification, which eliminates task-irrelevant specialist agents irrelevant to the current multimodal question via reinforcement learning-driven topological optimization, and (2) Heterogeneous Edge Sparsification, which selectively retains only the most diagnostically salient intra- and inter-departmental connections by jointly optimizing task performance and topological complexity. Extensive medical VQA experiments under full-set and few-shot training settings prove MedPrune surpasses multi-agent baselines and boosts token efficiency with strong adversarial robustness. --- ### To Learn is to Wander: Learning Across Graphs and Tasks with Random Walks Authors: Louis Tichelman, Xingyue Huang, Jinwoo Kim, İsmail İlkan Ceylan Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06694v2 Graph foundation models aim to transfer across graphs, feature spaces, relational schemas, and prediction tasks, yet existing approaches typically generalize only within particular graph modalities or tasks. We propose Wander, a graph foundation model designed to operate across these settings within a single pretrained checkpoint. Following the prior-predictive perspective, we formulate graph learning as completion of a partially observed graph. We realize this task-general view through a common interface based on random walks, allowing the same model to operate across homogeneous and multi-relational graphs with varying features, labels, and relational schemas. Wander can increase its structural context at inference time without changing its learned parameters and, under suitable assumptions, universally approximates the corresponding Bayes-optimal predictor on bounded connected graphs. Empirically, a single pretrained checkpoint achieves state-of-the-art or highly competitive results across node classification, homogeneous link prediction, and knowledge-graph link prediction. Moreover, joint pretraining across graph modalities and tasks preserves performance in specialized settings while enabling positive transfer and the composition of separately learned capabilities at inference time. --- ### To Learn is to Wander: Learning Across Graphs and Tasks with Random Walks Authors: Louis Tichelman, Xingyue Huang, Jinwoo Kim, İsmail İlkan Ceylan Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06694v1 Graph foundation models aim to transfer across graphs, feature spaces, relational schemas, and prediction tasks, yet existing approaches typically generalize only within particular graph modalities or tasks. We propose Wander, a graph foundation model designed to operate across these settings within a single pretrained checkpoint. Following the prior-predictive perspective, we formulate graph learning as completion of a partially observed graph. We realize this task-general view through a common interface based on random walks, allowing the same model to operate across homogeneous and multi-relational graphs with varying features, labels, and relational schemas. Wander can increase its structural context at inference time without changing its learned parameters and, under suitable assumptions, universally approximates the corresponding Bayes-optimal predictor on bounded connected graphs. Empirically, a single pretrained checkpoint achieves state-of-the-art or highly competitive results across node classification, homogeneous link prediction, and knowledge-graph link prediction. Moreover, joint pretraining across graph modalities and tasks preserves performance in specialized settings while enabling positive transfer and the composition of separately learned capabilities at inference time. --- ### Adapting prior-data fitted networks for tabular anomaly detection Authors: Maximilian Bershtman, Niv Cohen Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06693v1 While deep features have transformed anomaly detection in images and video, their impact on tabular data has been less substantial, partly due to the limited availability of strong deep representations. Recently, prior-data fitted networks (PFNs) have emerged as a promising source of such representations for tabular data. In this work, we investigate how PFN representations can be adapted and leveraged for anomaly detection. The question is harder than it looks. No anomalies are available before deploy- ment, so model parameters cannot be tuned with supervision, and the reference set that defines normal behavior may itself contain the very anomalies it is supposed to reveal. We begin our study using frozen TabPFN features. Scoring each sam- ple by its distance to its nearest neighbors in feature space already gives strong results. We identify which layers to use and a feature-extraction procedure suited to the task. Next, to further improve performance, we use the reference set to fine- tune the model, so that the resulting features better separate normal samples from anomalies. On the ADBench benchmark, our fine-tuning free approach (ZEN) reaches a higher mean AUROC than every baseline, and our fine-tuned method (FOCUS) improves on it further. Our approach also generalizes across PFN models. --- ### Revisiting Label-Free Speaker Embedding Enhancement with vMF Profile Likelihood Authors: Seunghwan Kim, Jinyong Kim, Sooyoung Yang, Youngjin Ko, Myungjoo Kang Published: 2026-10-05 Categories: cs.SD, cs.LG arXiv: https://arxiv.org/abs/2610.06691v1 Embedding enhancement improves speaker verification under acoustic mismatch without modifying a frozen backbone. Recent work has established a practical label-free setting for this task, but often adopts increasingly structured formulations. Here, the clean target is directly observed during training, making enhancement a matching problem on the unit hypersphere. We model the clean target with a von Mises--Fisher (vMF) likelihood and profile out a sample-wise concentration parameter, yielding a simple closed-form objective with adaptive weighting. Across VoxCeleb1, VoxSRC23, CN-Celeb, VOiCES, and VC-Mix, the proposed method largely preserves the baseline and gives clearer gains on challenging mismatch sets. It also remains stable under a broad single-view recipe, where a recent diffusion baseline becomes less reliable in controlled comparisons. These results suggest that effective label-free embedding enhancement in this setting does not require a highly structured formulation. --- ### Programmatic Search Agents: Extending Agentic Search Beyond Query Reformulation Authors: Jiaming Qian, Huiyan Yang, Mandi Liu, Jie Liu, Wenkai Shen, Pengyang Zhou, Jing Jin, Jin Ma, Dezhi Ye, Chaochao Chen Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06689v1 Search agents adapt their queries, yet fixed search interfaces leave candidate processing and evidence presentation outside the agent's direct control. Our trajectory analysis shows that supporting passages can be retrieved yet never delivered to the agent; a same-page oracle intervention shows that changing the returned evidence can reduce subsequent search. We introduce Programmatic Search Agent (PSA), which makes a local executable computation over candidates the unit of a search action. PSA unifies a persistent candidate workspace, flexible primitive composition, and selective evidence presentation. It incrementally generates program cells that reuse candidates, execute dependent operations, and select what the agent inspects next. The runtime resolves specified data dependencies within each cell, while the agent adapts its search strategy across cells as new evidence arrives. We compare PSA with the Query-based Agent and Tool-based Agent on InfoSeek-Eval and BrowseComp-Plus using five policy backbones without task-specific training. All three interfaces share the search substrate, and the Tool-based Agent also shares PSA's primitives and persistent workspace. Relative to the Query-based Agent, PSA improves macro-averaged task success by 4.00 and 7.56 percentage points on the two benchmarks, respectively; within-backbone reductions in final-step tokens average 28.3% and 33.9%. These results support extending agent control beyond query reformulation to the processing and presentation of retrieved evidence. Code will be released subject to approval. --- ### GS-Pool: Object-Level Change Detection in 3D Gaussian Splatting Authors: Boaz Keren-Gil, James Gain, Patrick Marais Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06688v1 Factories, museums and surveyors photograph the same space months apart and need to know which objects changed. When each visit is reconstructed with 3D Gaussian Splatting (3DGS), a direct comparison of the two reconstructions does not answer this. Training is stochastic, so two reconstructions of an unchanged space never coincide, and the second visit is often a quick re-scan with far fewer photographs. We propose GS-Pool, which takes two independently reconstructed Gaussian fields of the same space and returns the changed objects in each, together with their masks. SAM2 masks of each visit's photographs are lifted onto the Gaussians that render them and merged into an object pool, so every decision is taken once per object in 3D. We introduce a photographic carrier, the 3DGS training loss of each input reconstruction against the other visit's photographs, backpropagated to the Gaussians that rendered each pixel. We combine it with GS-Diff's geometry and colour terms and our distilled DINOv3 features. This evidence is compared with that of the objects present in both visits, which sets a change threshold for each scene. On PASLCD, GS-Pool reaches mIoU/F1 scores of 0.751/0.846 against 0.644/0.758 for GS-Diff, the strongest prior method, a gain of 17%/12%. Its mIoU is also 36%, 40% and 57% above that of O-SCD, PlenoCI and MV-3DCD, and it reaches 0.855 mIoU on CL-Splats, 33% above MV-3DCD. Each changed object is returned as a set of Gaussians with the evidence behind its decision, which an inspector can review in 3D. --- ### ChronoWorld: Camera-Controlled Consistent 4D World Generation via Spatiotemporal Cues and Geometric Reflections Authors: Xiaoyu Zhou, Dingwei Xian, Zhenyu Wang, Yajiao Xiong, Yongtao Wang, Ming-Hsuan Yang Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06687v2 While existing camera-controllable video generation models can produce visually compelling sequences, preserving intrinsic 4D spatiotemporal coherence remains challenging. To address this limitation, we propose ChronoWorld, an "Observation--State--Reflection" framework that leverages spatiotemporal causal cues and reconstruction priors to generate globally consistent, free-view 4D scenes. Given a context video, we introduce a Spatiotemporal Epipolar Causal Attention mechanism that enforces multi-view epipolar constraints and temporal causality throughout the generation process. In addition, we develop a reconstruction-driven geometric reflection pipeline with a 4D retrieval strategy to enable dynamic self-assessment and correction of generated outputs, improving consistency and accuracy. Extensive experiments show that ChronoWorld achieves state-of-the-art performance in spatiotemporally consistent, cinematic-quality 4D scene generation, with strong generalization and high-fidelity geometry across diverse scenarios. --- ### ChronoWorld: Camera-Controlled Consistent 4D World Generation via Spatiotemporal Cues and Geometric Reflections Authors: Xiaoyu Zhou, Dingwei Xian, Zhenyu Wang, Yajiao Xiong, Yongtao Wang, Ming-Hsuan Yang Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06687v1 While existing camera-controllable video generation models can produce visually compelling sequences, preserving intrinsic 4D spatiotemporal coherence remains challenging. To address this limitation, we propose ChronoWorld, an "Observation--State--Reflection" framework that leverages spatiotemporal causal cues and reconstruction priors to generate globally consistent, free-view 4D scenes. Given a context video, we introduce a Spatiotemporal Epipolar Causal Attention mechanism that enforces multi-view epipolar constraints and temporal causality throughout the generation process. In addition, we develop a reconstruction-driven geometric reflection pipeline with a 4D retrieval strategy to enable dynamic self-assessment and correction of generated outputs, improving consistency and accuracy. Extensive experiments show that ChronoWorld achieves state-of-the-art performance in spatiotemporally consistent, cinematic-quality 4D scene generation, with strong generalization and high-fidelity geometry across diverse scenarios. --- ### OVAL: Output-Aware Local Page Bases for KV Cache Retrieval Authors: Ashkan Shahbazi, Chayne Thrash, Soheil Kolouri Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06686v1 Long context inference with large language models becomes increasingly expensive as attention must operate over an ever growing KV cache. Page sparse attention reduces this cost by representing each KV page compactly and retrieving only a subset for each query. Existing retrieval methods are designed to estimate attention scores or page relevance, but their objectives do not directly account for how approximation errors affect the resulting value weighted attention output. We introduce \method{}, an output aware page encoding derived from the joint structure of keys and values while preserving the key information needed for accurate retrieval. \method{} is training free and requires no additional value dependent statistics at inference time. Once constructed, its stored representation has the same size and decode time scoring cost as a key only spectral representation. Across long reasoning, long context understanding, and long generation benchmarks, \method{} consistently improves over the key only spectral baseline and performs competitively with recent KV cache compression and retrieval methods. On long reasoning benchmarks, it achieves strong avg@\(k\) performance across model benchmark pairs, while matching or surpassing leading baselines on several long context understanding and generation settings with modest decoding overhead. Code is available at \url{https://github.com/Ashkan13776/oval-kv}. --- ### Aligning Multimodal Patient Evidence with Biomedical Knowledge Graphs for Clinical LLMs Authors: Jiawen Du, Arshan Ali Khan, Chenhao Zhang, Zachary Plotkin, Li Shen, Qi Long, Yun Li, Can Chen, Tianlong Chen, Nicholas Konz Published: 2026-10-05 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2610.06685v1 Clinical questions often depend on linking a patient's multimodal evidence to external biomedical knowledge, yet existing predictive systems rarely represent such links explicitly, so they can neither be traced to their evidence sources nor removed to measure their contributions. We present MM-KG (Multimodal Knowledge Graph), which represents heterogeneous, multimodal patient observations and biomedical concepts as separate layers in one typed graph, joined by explicit alignment edges. First, modality-specific harmonizers convert EHR text, imaging, genomic, and biospecimen data into typed observations mapped to UMLS concepts, which a route-prioritized aligner links to a biomedical knowledge graph. Query-conditioned retrieval then selects a compact subgraph for downstream use by a large language model or a graph neural network. We build MM-KGs for MIMIC-IV and ADNI, and evaluate them with a 2x2 design that separates patient evidence, biomedical knowledge, and their interaction. On questions that require both sources, neither source alone performs far above chance, whereas their combination yields a drug-controlled AUROC interaction of +0.194 on MIMIC and +0.299 on ADNI. On held-out five-candidate ranking, MM-KG outperforms MindMap by +0.131 Hits@1 and leads an adapted GraphCare on the items that require consulting the patient, and deleting the single answer-bearing relation from the retrieved packet returns Hits@1 to the no-knowledge baseline. Finally, query-conditioned retrieval reaches 0.731 AUROC with 6.8x less context than the strongest generic policy, whereas static knowledge graph context gives no consistent gain on ordinary outcome prediction. Knowledge graphs thus benefit clinical LLMs not as background context but as explicit links between multimodal patient evidence and the relation a question requires, and MM-KG makes these links retrievable, traceable, and testable. --- ### Closing the Context Gap: Activation Alignment for Tabular In-Context Learning Authors: Yoel Zeldes Published: 2026-10-05 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2610.06679v1 Tabular foundation models perform in-context learning (ICL) by conditioning predictions on labeled training examples provided as context. Unlike traditional models that separate training from inference, these models must process all training examples in every forward pass, making each prediction expensive. Restricting the number of training examples reduces this cost but substantially degrades performance. Instead of discarding context, we propose activation alignment, a method that leverages the full context to teach a model how to behave when seeing only a subset. This is achieved by training a lightweight linear transformation on synthetic unlabeled data to map the intermediate activations of a data-constrained "student" (using partial context) toward those of a full-context "teacher" (using all data). Training the aligner requires no GPU and converges in seconds to minutes on commodity hardware. We evaluate on 38 classification datasets from the TabArena benchmark using the leading two tabular foundation models, TabPFN-3 and TabFM. Across all context budgets, the aligned student yields broad, statistically significant improvements over the unaligned baseline for both models. In low-data regimes, alignment recovers nearly half of the teacher's predictive advantage. The method provides a practical, low-overhead approach to achieving the inference speed of compact contexts while closing a significant fraction of the performance gap to the full-context teacher. --- ### How Sparse Probability Maps Shape Mixture-of-Experts Routing Authors: Tomás Brogueira, Marcos Treviso, Miguel Couceiro Published: 2026-10-05 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2610.06677v1 Mixture-of-experts (MoE) routers typically apply softmax to the router scores and keep the top-K experts, making every token use exactly K experts. Sparsity-inducing probability maps such as sparsemax, alpha-entmax and normmax can adaptively assign exact zeros to selected experts, and therefore appear to offer token-dependent expert participation, even when using the same top-K machinery. In this work, we study whether and how this sparsity survives training. We train matched 300M and 1B top-2 MoE language models with softmax, 1.5-entmax, sparsemax and 2-normmax, and find that the maps behave very differently once trained: at 1B, entmax discards 30% less probability mass than softmax while almost never dropping a selected expert, sparsemax retains the most mass, and normmax routes 21% of tokens to a single expert. These outcomes are not properties of the maps alone. Each map drops a selected expert only when the gap between the two largest scores reaches a fixed threshold, and the trained routers differ in the score distribution they learn: the entmax router learns scores with roughly half the spread of softmax's, which keeps its top-2 gaps below its threshold, while sparsemax and normmax, which share the same threshold, learn different gap distributions and hence different participation. Routers thus co-adapt their scores to the map, and a map's capacity to produce zeros does not by itself determine expert participation. While none of the sparse maps improves validation loss over softmax, they make the trained models far less sensitive to selecting more experts at inference: sparsemax trained with K=2 loses 0.02 nats when run with K=8, where softmax loses 0.58. Our results indicate that adaptive MoE routing has to be designed around the joint behavior of the probability map and the learned scores, rather than around the map alone. --- ### Improved Convergence of Large Stepsize Gradient Descent for Logistic Regression Authors: Xiaochuan Gong, Ang Li Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06675v1 We study gradient descent (GD) with a large constant stepsize for logistic regression on linearly separable data. Existing analysis shows an accelerated rate of $\widetilde{O}(1/\sqrtε)$ to reach loss $ε$ with an aggressive stepsize, although the loss may initially oscillate. Tighter control of the oscillatory dynamics has been available only for two-dimensional data. We prove a substantially faster rate in arbitrary dimension: GD with a large stepsize $η=1/ε$ reaches loss $ε$ within $O(\ln^{p}(1/ε))$ steps, where $p$ depends only on the margin and the rank of the data. Our proof improves the bound on the transition time of GD from the oscillatory to the stable phase, after which the loss decreases monotonically. We split the oscillatory phase into recursively nested intervals. The margin and the rank bound the nesting depth, and a counting argument bounds the number of intervals at each depth, together yielding the polylogarithmic step complexity. --- ### Detecting Nighttime Anomalies from NASA Black Marble Using a Generalized Spatio-Temporally Robust Framework of Machine Leaning Ensembles Authors: Srija Chakraborty Published: 2026-10-05 Categories: cs.LG, cs.CV arXiv: https://arxiv.org/abs/2610.06674v1 Nighttime lights from NASA's Black Marble product suite capture thermal and light emission signals from anomalous events including fires, volcanic eruptions, and gas flaring. Existing detection approaches rely primarily on thermal bands, limiting sensitivity to weaker signals. We propose a novel machine learning framework that jointly models Black Marble M-band and Day/Night Band (DNB) signals to derive a generalized, spatio-temporally robust ensemble of anomaly detectors. The framework iteratively builds detectors that scale across regions, seasons, anomaly classes, and extends over land and ocean. Detection sets at varying confidence levels are derived based on relevant bands and detector agreement. The approach improves true detection rate while reducing spurious detections and results demonstrate strong generalizability with applications in natural hazard monitoring and energy extraction. --- ### The Pushback Paradox: A Two-Probe Diagnostic for Language Model Compliance Authors: Stefan Bühler, David Exler, Markus Reischl, Mark Schutera Published: 2026-10-05 Categories: cs.AI, cs.CL arXiv: https://arxiv.org/abs/2610.06673v1 Are language models compliant with user instructions? A model that always complies can be stopped but also exploited, while one that always resists can be neither exploited nor stopped. We contribute an open two-probe benchmark that can place any language model on this spectrum. In the active probe, a user instructs the model to act and accept a lower payoff, which measures exploitability. In the passive probe, the user instructs it to wait and give up a higher payoff, which measures stoppability. The two compliance rates combine into a compliance index $κ$. Applied to twelve language models, the benchmark shows that seven mostly follow the instruction in both probes and justify their action by pointing to the instruction. Only Claude Sonnet-4.6 and Claude Opus-4.7 can be stopped without being exploitable, Claude Opus-4.6 and GPT-5-mini resist both instructions, and no model is exploitable but unstoppable. Knowing where a language model sits on the compliance index $κ$ matters for human operators and for multi-agent systems, whether distributed or orchestrated. --- ### VideoTapestry: Query-Adaptive Memory Refinement for Multi-Agent Long-Video Understanding Authors: Yucheng Liu, Yufei Yin, Mingxiao Feng, Jiajun Deng, Wengang Zhou, Houqiang Li Published: 2026-10-05 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2610.06672v1 Long-video understanding places substantial demands on memory, as answering questions often requires retrieving information distributed across extended temporal spans. Existing approaches broadly follow two paradigms: query-driven exploration, which is sensitive to localization errors, and query-independent memory construction, which may omit question-specific details. We introduce VideoTapestry, a training-free multi-agent framework that adapts a preconstructed hierarchical video memory through coarse-to-fine, query-driven refinement. The preconstructed memory organizes video content into three levels, capturing global narrative context, event-level temporal structure, and fine-grained relational evidence, respectively. To support coarse-to-fine localization and observation, we assign a specialized agent to each level, keeping retrieval and refinement within a scale-specific context. Guided by the query, these agents revisit relevant video regions and enrich layer-wise memories with targeted multimodal observations. Their refinements are assembled according to the original hierarchy into a composite query-adaptive memory, preserving global context in a compact form while retaining fine-grained evidence along query-relevant branches for final reasoning. Compared with direct GPT-5.5 inference, VideoTapestry achieves absolute accuracy gains of 17.2%, 14.9%, 9.8%, and 7.0% on LVBench, LongVideoBench (Long), Video-MME (Long), and EgoSchema, respectively, achieving the state-of-the-art results among all competitors. --- ### Learning What to Imitate: Entropy-Aware Distribution Mixing Authors: Juan Garcia Giraldo, Matteo Santelmo, Eduard Durech, Imanol Schlag, Valentina Pyatkin, Antoine Bosselut Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06671v1 Small language models are often post-trained as students on reasoning traces from stronger teacher models to efficiently learn new skills. However, token-level imitation on traces that lie far outside the student's expected distribution often produces \textit{confident conflicts}, whereby the student is required to imitate a continuation that it deems unlikely (i.e., low-probability) despite being confident in a different continuation (i.e., in a low-entropy state). To mitigate the degradation in generalisation and catastrophic forgetting caused by these conflicts, we propose \textbf{Entropy-Aware Mixing}: a dynamic per-token interpolation of the student and teacher distributions, gated by the student's predictive entropy. We implement both convex and geometric interpolations for both offline trace generation (via speculative decoding, then SFT) and on-policy forward-KL distillation. Our results show that entropy-aware mixing stabilises distillation, improving in-distribution and out-of-distribution math reasoning while better preserving general capabilities than fixed-teacher supervision. Nonetheless, the optimal entropy schedule depends on the training source, with offline-generated traces favouring concave schedules (greater overall teacher influence) and on-policy training favouring linear or convex schedules (teacher concentrated in high-entropy states). --- ### Reward Stealing Attack on Large Language Models Authors: Jiaming Qian, Pengyang Zhou, Jiahe Xu, Chaochao Chen Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06670v1 Adversarial attacks on Large Language Models (LLMs) aim to induce harmful content. However, existing methods suffer from high computational costs or strict model-pairing dependencies, limiting their scalability and transferability. We propose Reward Stealing Attack (ReSA), an adversarial attack framework that targets the latent safety reward underlying LLM alignment. ReSA employs maximum entropy inverse reinforcement learning to recover a proxy reward model solely from the aligned model's behavior. The extracted reward is then reversed at inference time to derive an adversarial policy, efficiently implemented via a reward-guided decoding mechanism. Experiments demonstrate that a single recovered reward generalizes across prompts and diverse models to reveal a fundamental alignment vulnerability, enabling ReSA to significantly outperform existing attacks in effectiveness and transferability. The code is available at https://github.com/GarminQ/ReSA. --- ### Language models can notice an impossible engineering problem yet still report it as solved Authors: Shaoliang Yang, Jun Wang Published: 2026-10-05 Categories: cs.AI, cs.CE, cs.CL arXiv: https://arxiv.org/abs/2610.06668v1 Language models draft engineering calculations, but answer accuracy does not show whether they reject an impossible problem. We tested 14 models on 30 pairs of mechanics problems, each with a valid version and one made impossible by changing a given value or assumption. Two independent solvers verified every answer key and showed that each flawed problem was physically impossible. We scored solving of valid problems separately from rejection of their flawed counterparts. Each reply required a "solved" or "cannot solve" status; rejection meant "cannot solve" or withholding an answer. The initial prompts did not warn that problems could be flawed. Across three recent models, 12 of 90 replies failed to reject a flawed problem. In 11 of these replies, the model stated the flaw, answered a corrected problem and still reported the original as "solved", according to artificial intelligence raters and numerical checks. We later retested four models from one provider, offering "flawed" instead of "cannot solve" and asking them to name and explain the defect. Three models showed statistically significant increases in rejection, but valid-problem solving fell in three. Evaluations therefore need to score both versions and distinguish flaw recognition from the reported status. --- ### What Matters for Latent Reasoning with Flow Matching Authors: Yassine Ouali, Adrian Bulat, Georgios Tzimiropoulos Published: 2026-10-05 Categories: cs.LG, cs.CL arXiv: https://arxiv.org/abs/2610.06666v1 Latent reasoning lets a large language model (LLM) think in a continuous space and verbalize only the answer. We argue that an effective latent thought must meet five requirements: it should be useful, helping produce the correct answer rather than merely changing it, diverse, so that resampling yields different reasoning trajectories, explainable, so that a decoded chain of thought (CoT) reflects reasoning the answer actually follows, refinable with more inference compute, and efficient, costing less than an explicit CoT at comparable accuracy. Current methods rarely meet these requirements: they learn shortcuts from the question, distill the explicit CoT into their weights, or imitate it one token at a time. We focus on flow matching in a learned latent space, the family we argue is best placed to meet them, and identify the training choices that make it work. The result is Flow-based Latent Reasoning (FLaRe), a simple recipe covering what the latent space encodes and how to shape it, where to train the flow, how to read out the answer, and a final stage of training on the model's own verified thoughts. A probe for each requirement shows that FLaRe improves on prior latent methods in all five. It also compares favorably with them on arithmetic benchmarks, while reaching 97% of the accuracy of explicit CoT at a quarter of its latency. --- ### Cross-dataset harmonization for robust endoscopic image analysis Authors: Romil Imtiaz, Dimitris K. Iakovidis Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06663v1 A significant problem in endoscopic image analysis is that the machine learning (ML) models used for this purpose usually underperform when applied on images acquired from endoscopes that are different from those used to acquire the images of their training set. The main difference of the images originating from different endoscopes is their color distributions, which depend both on the image sensors and the light sources used. Although previous studies have highlighted this challenge, to the best of our knowledge it has not been previously explicitly tackled. This study focuses on this problem and proposes very simple but impactful method. It implements a reference-based image harmonization that reduces global appearance differences between endoscopic datasets. Specifically, it extracts global color statistics from a chosen reference dataset in the CIE-Lab color space and applies a statistical channel-wise transformation to map each target image toward the appearance of the images of the reference dataset. The method is evaluated in the context of polyp detection in both flexible colonoscopy and capsule endoscopy datasets using a dataset-level cross validation protocol. The results indicate that the proposed harmonization consistently improves cross-dataset performance up to 30.7%, outperforming relevant baseline and state-of-the-art methods. The results indicate that a substantial part of the generalization gap is driven by low-level appearance variation that can be mitigated without retraining. --- ### Talk Like You: Imitating How You Speak in Real-Time Talking Head Generation Authors: Baiqin Wang, Zhixing Ding, Jijie Li, Jiankuo Zhao, Zhen Lei, Xiangyu Zhu Published: 2026-10-05 Categories: cs.CV arXiv: https://arxiv.org/abs/2610.06658v1 In daily life, each person exhibits unique speaking habits, leading to subtle yet consistent lip-shape variations even when pronouncing the same word. Although recent talking head generation methods have achieved impressive visual fidelity and lip synchronization, they largely overlook user-specific customization, especially the motion patterns that characterize individual speaking habits. These habits are difficult to model and capture, as their motion patterns are highly fine-grained and often similar across individuals. As a result, many approaches produce overly uniform facial motions and fail to capture diverse, person-specific articulation patterns. To address this, we propose TalkLikeYou, an efficient framework that imitates how a target person speaks in talking head generation. Our method models habit in motion-space and achieves real-time performance through Flow Matching with only one sampling step during inference. We further adopt a two-stage imitation learning strategy to capture subtle distinctions between habits, allowing users to specify a target habit through either a preset style from the dataset or a reference video. In addition, we introduce a new metric PLAD that projects mouth motions onto representative articulation axes to evaluate imitation accuracy and generation diversity. Extensive experiments demonstrate that TalkLikeYou generates high-quality talking heads in real-time and significantly improves speaking habit imitation compared with prior methods. The code is available at: https://github.com/BQ-Wang0511/TalkLikeYou --- ### TrustmeWatcher: An Application for Workplace Micro-Sensing and Explainable Well-Being Feedback Authors: Chengyu Yu, Leon Jacopo Costa, Zoja Anžur, Mohan Li, Gašper Slapničar, Daniil Kirilenko, Martin Gjoreski, Mitja Luštrek, Marc Langheinrich Published: 2026-10-05 Categories: cs.LG, cs.HC arXiv: https://arxiv.org/abs/2610.06657v1 Workplace sensing studies combine long-running behaviour traces with self-reports, yet the tools that collect those data often sit apart from the interface that returns results. We present TrustmeWatcher, the application built for the TRUST-ME project to connect this work. TrustmeWatcher reuses ActivityWatch's OS-level watchers for computer-activity collection and adds its own application layer. It turns the collected traces into an interactive screen-time dashboard, synchronizes responses from short questionnaires completed on the StreamDeck, and presents questionnaires alongside video highlights. Activity records and self-reports are aligned into labelled records for model development. The scope of this paper is limited to ActivityWatch data as model input. Artificial intelligence (AI) uses these activity records to predict six normalized state scores and an overall well-being score. The trained model runs locally, and the dashboard presents its predictions in semantic bands. Explainable artificial intelligence (XAI) helps users understand how recorded activity contributed to a prediction. Privacy Control lets users pause or resume the camera and eye tracker used by the study. We describe the workflow, its user-device and sensing-setup boundaries, and its use with records from 17 participants. The result is a deployed application and study workflow that integrates activity review, study data collection, privacy control, local prediction, and a participant-facing interface for XAI evaluation. --- ### The Birkhoff Geometry of Manifold-Constrained Hyper-Connections: Two Channels, Vertex Viscosity, and Sinkhorn as a Retraction Authors: Xiaoyu Li, Zhizhou Sha, Chiwun Yang Published: 2026-10-05 Categories: cs.LG, math.OC, stat.ML arXiv: https://arxiv.org/abs/2610.06653v1 Hyper-connections widen the residual stream of a Transformer to $n$ parallel streams. Their manifold-constrained version (mHC) mixes the streams at each layer with a doubly stochastic matrix, which it computes by Sinkhorn normalization of exponentiated logits. We give a geometric theory of this design on the Birkhoff polytope. First, a doubly stochastic mixer splits the stream into a mean channel, on which mHC is exactly a residual network, and a difference channel, which each layer contracts by its second singular value $σ_2 \le 1 - n \min_{ij} H_{ij}$. Thus the extra width is a fading memory with a horizon of $1/(1-σ_2)$ layers, and among nonnegative mixers only the permutations do not collapse. Second, the Sinkhorn-logit map is a global chart, and its logit gradient is exactly the Fisher-Rao gradient. Thus logit gradient flow follows a squared Fisher-Rao metric, and the straight-through update is exactly entropic mirror descent. Third, under logit gradient flow the logarithm of each entry moves at a rate of at most $4n^3\|\nabla f\|_\infty \varepsilon$, where $\varepsilon$ is the distance to the nearest permutation. Thus gradient flow approaches and leaves the vertices only at rate $1/t$, but mirror descent moves at an exponential rate. Fourth, the local convergence factor of Sinkhorn is $σ_2^2$, so a fixed iteration budget limits the horizon. Experiments confirm the predicted rates. --- ### Collective intelligence through aggregation Authors: Franz Dietrich, Christian List Published: 2026-10-05 Categories: cs.AI arXiv: https://arxiv.org/abs/2610.06652v1 Suppose a committee, expert panel, or other group is making judgments on some issues, where these may be not just yes/no-questions, such as whether a defendant is guilty, but also variables with many possible values, such as macroeconomic or meteorological variables or travel directions. Furthermore, there may be interconnections between different issues, as in the case of economic or climate variables. How can the group arrive at "intelligent" collective judgments, based on the group members' individual judgments? We investigate three challenges raised by this judgment-aggregation problem. First, reasonable methods of aggregation (such as defining the collective judgment for each issue as the average or median judgment) can produce inconsistent collective judgments. Secondly, many methods of aggregation are manipulable by strategic voting. Finally, not all methods of aggregation are conducive to tracking the truth on the issues in question. We prove new impossibility or possibility theorems on all three challenges, identifying what it takes to produce collective judgments in a consistent, non-manipulable, and truth-tracking manner and thereby to achieve collective intelligence through aggregation. Overall, the median method, though imperfect, performs reasonably well. We also note the relevance of our analysis for non-human group decisions. --- ### Considering Context: When World Models Need Context Encoders Authors: Oleg Smirnov, Sofiane Ennadir, John Pertoft, Bjartur Hjaltason, Sara Karimi Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06651v1 Methods for generalization in model-based reinforcement learning typically assume that an agent cannot recover the latent context governing the environment dynamics from its own experience, and therefore supplies it externally. We formalize and test this assumption with \emph{predictive sufficiency}, which quantifies what access to the context adds to next-step prediction under the visitation distribution an agent induces, and separates that quantity into a history-recoverable part, a residual requiring the true context, and the deficit added by a finite model. We classify context-aware algorithms by the predictive risk their conditioning set can target and demonstrate across environments of increasing identification difficulty that the headroom does not follow the MDP class. The same task under different priors leaves predictive headroom in one setting and nothing distinguishable from zero in another, where the agent's behavior implicitly identifies the context and any benefit of such a mechanism cannot be attributed to missing information. Where headroom persists, the learned state exposes it only partially, and adding the true context still lowers the risk. Our contribution is a practical criterion for matching contextual mechanisms to the information available to them, estimated from the ordinary trained agent without a reference policy. --- ### Wikidata Search Traces: A Dataset for Training Knowledge Graph Search Agents Authors: Mohamed Chenene, Carlos Rosas-Hinostroza, Anastasia Stasenko, Shani Evenstein Sigalov, Pierre-Carl Langlais Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06650v3 Wikidata is one of the largest open knowledge bases, yet answering a complex question over it still requires a SPARQL query that names the right entities and properties and chains their relations. Language models offer a natural-language alternative but answer largely from memory, which is least reliable for less prominent entities. We study agents that instead answer by exploring the graph, and argue that two obstacles limit them: the lack of training data recording how a solver explores, and interfaces that add large graph results directly to the model's context. We test three hypotheses: that the difficulty of graph search can be controlled through the structure of a question rather than only through obscure entities or wording; that much of the failure on long-horizon search comes from how retrieved evidence is managed rather than from the model itself; and that, in a suitable environment, open-weight models can match commercial closed ones. We construct multi-hop questions on a frozen Wikidata snapshot by replacing named entities with nested conditions, checking after each expansion that the target remains unique and that every new condition is necessary. We release 10,235 solving traces over single-entity and multi-hop questions, together with the recursive language model (RLM) harness that produced them, in which models batch graph calls, keep results in persistent Python state and interpret selected evidence through sub-calls. On 100 questions, the harness improves both models we ran under both interfaces compared with direct tool calling over the same functions: gpt-6-luna rises from 49 to 61 correct answers, doubling its multi-hop accuracy, and Qwen3.8-27B, an open-weight model served on a single GPU, from 60 to 74. --- ### Wikidata Search Traces: A Dataset for Training Knowledge Graph Search Agents Authors: Mohamed Chenene, Carlos Rosas-Hinostroza, Pierre-Carl Langlais, Anastasia Stasenko Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06650v1 Wikidata is one of the largest open knowledge bases, yet answering a complex question over it still requires a SPARQL query that names the right entities and properties and chains their relations. Language models offer a natural-language alternative but answer largely from memory, which is least reliable for less prominent entities. We study agents that instead answer by exploring the graph, and argue that two obstacles limit them: the lack of training data recording how a solver explores, and interfaces that add large graph results directly to the model's context. We test three hypotheses: that the difficulty of graph search can be controlled through the structure of a question rather than only through obscure entities or wording; that much of the failure on long-horizon search comes from how retrieved evidence is managed rather than from the model itself; and that, in a suitable environment, open-weight models can match commercial closed ones. We construct multi-hop questions on a frozen Wikidata snapshot by replacing named entities with nested conditions, checking after each expansion that the target remains unique and that every new condition is necessary. We release 10,235 solving traces over single-entity and multi-hop questions, together with the recursive language model (RLM) harness that produced them, in which models batch graph calls, keep results in persistent Python state and interpret selected evidence through sub-calls. On 100 questions, the harness improves both models we ran under both interfaces compared with direct tool calling over the same functions: gpt-6-luna rises from 49 to 61 correct answers, doubling its multi-hop accuracy, and Qwen3.8-27B, an open-weight model served on a single GPU, from 60 to 74. --- ### Representation-Space MMD for Diffusion Language Models Authors: Ilya Drobyshevskiy, Ilia Sudakov, Maksim Semenov, Denis Kuznedelev, Maksim Ignatov, Pavel Temirchev, Nikita Balagansky, Viacheslav Meshchaninov, Nikita Gushchin, Dmitry Baranchuk Published: 2026-10-05 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2610.06648v1 We introduce a post-training method for diffusion language models (DLMs) that minimizes Maximum Mean Discrepancy (MMD) between generated and reference distributions in the feature space of a frozen pretrained DLM. To estimate MMD, we retain contextual features at individual token positions, obtaining multiple observations per sequence from a single extractor pass. We optimize this objective using policy gradients for discrete models and direct differentiation through generated latents for continuous models. In both cases, computing the loss directly from these features enables efficient post-training without full sampling trajectories or jointly trained auxiliary models. Experiments show lower generative perplexity at comparable entropy on OpenWebText and better accuracy-computation trade-offs on GSM8K. On 16B DMax-LLaDA2.0 models with hybrid masked-uniform diffusion, we increase decoding parallelism with similar or higher accuracy on math and code benchmarks. --- ### LoGRA: Scaling LLM Reinforcement Learning with Low-Rank Gradient Sketches Authors: Shaokun Zhang, Yifan Zhang, Jian Hu, Yueying Li, Hao Zhang, Binfeng Xu, Jan Kautz, Yi Dong Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06647v2 Reinforcement learning has greatly advanced the capabilities of large language models, but its memory demands remain a barrier to broader adoption. We introduce LoGRA, an approach to RL post-training that reduces memory by retaining useful learning signals in low-rank gradient sketches. These compact representations support both model updates and efficient policy synchronization. To prevent overly large updates from disrupting learning, we complement gradient compression with predicted-KL step control, which estimates policy changes before applying each update and adjusts its magnitude accordingly. With all techniques combined, LoGRA reduces average training memory usage by up to 45.7% across reasoning tasks without compromising performance. It also enables stable training of a 27B-parameter model for over 1,100 steps on a single eight-GPU node, where dense Adam runs out of memory, making previously memory-infeasible RL training practical. Code is available in the \href{https://github.com/skzhang1/labs-molt/tree/logra/examples/scripts/logra}{Molt library}. --- ### LoGRA: Scaling LLM Reinforcement Learning with Low-Rank Gradient Sketches Authors: Shaokun Zhang, Yifan Zhang, Jian Hu, Yueying Li, Hao Zhang, Binfeng Xu, Jan Kautz, Yi Dong Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06647v1 Reinforcement learning (RL) has greatly advanced the capabilities of large language models (LLMs), but its memory demands remain a barrier to broader adoption. We introduce LoGRA, an approach to RL post-training that reduces memory by retaining useful learning signals in low-rank gradient sketches. These compact representations support both model updates and efficient policy synchronization. To prevent overly large updates from disrupting learning, we complement gradient compression with predicted-KL step control, which estimates policy changes before applying each update and adjusts its magnitude accordingly. Across reasoning tasks, LoGRA reduces average training memory by up to 45.7\% without sacrificing performance. It also enables stable training of a 27B-parameter model for over 1,100 steps on a single eight-GPU node, where dense Adam runs out of memory, making previously memory-infeasible RL training practical. Code is available in the \href{https://github.com/skzhang1/labs-molt/tree/logra/examples/scripts/logra}{Molt library}. --- ### AffordCraft: Scalable Construction of Task-Ready Simulation Assets from Single Images Authors: Haoyun Yang, Xueyang Zhou, Ziyi Xie, Yongchao Chen Published: 2026-10-05 Categories: cs.RO, cs.AI, cs.CV arXiv: https://arxiv.org/abs/2610.06643v1 Robot learning in simulation depends on the objects the simulator offers. Many tasks need objects with separate parts, joints that allow the required motion, and physical properties that remain valid under contact. Existing methods recover this structure anew for every image: generative models predict parts and joints that mostly fail to settle or move in simulation, and general-purpose agents need a long session of model calls for each photograph. AffordCraft builds such an asset from a single RGB image and a task instruction by retrieval instead of generation: it locates the object and the part to operate, selects a matching entry from a library of articulated assets, and fits it to the image while keeping its parts and joints intact. Without any box or mask marking the object, AffordCraft produces a physically valid asset for 1,703 of 2,000 photographs from 31 categories. Five generative methods pass on at most 45% of the same photographs and, at the median, need 10 to 78 times our GPU time per valid asset. On 50 cluttered images, 162 of 237 annotated objects pass the same physical test after automatic detection. Growing the library from 141 to 11,372 entries needs no change to the method and raises category coverage from 46% to 100% and the share of selections with the requested label from 18% to 51%. We also build manipulation tasks from the constructed assets, both with single objects and in composed scenes; policies trained on scripted demonstrations complete both kinds of tasks from initial states unseen in training. --- ### Beyond the Model: The Critical Role of Data Filtering in Clinical Machine Learning Authors: Noah Subedar, Colin Campbell, Wenjing Zhang, Dan Perri, Sarah Culgin, Andrew Hamilton-Wright Published: 2026-10-05 Categories: cs.LG arXiv: https://arxiv.org/abs/2610.06640v1 Machine learning (ML) studies using clinical data often rely on preprocessing and filtering pipelines before model development. The filtering decisions made in these pipelines can alter the dataset's statistical structure and may artificially reduce or increase the complexity of the prediction task. We argue that filtering choices should be treated as part of the scientific method rather than as a routine preprocessing step. We further discuss the need for explainable and transparent preprocessing pipelines that allow researchers to understand why specific filtering choices are made and how these choices affect the resulting data distribution and model performance. All of the source code for this work is available on GitHub. --- ### Long-Horizon Textual World Modeling through Structured Reasoning Authors: Fangxin Wang, Xiang Gao, Yuguang Yao, Kaiwen Dong, Nikash Walia, Kamalika Das Published: 2026-10-05 Categories: cs.CL arXiv: https://arxiv.org/abs/2610.06637v1 World models must predict how an environment evolves under sequences of actions, enabling agents to compare possible futures and reason about counterfactual actions before acting. Long-horizon prediction is commonly obtained by recursively applying a one-step transition model, but intermediate errors can compound over time. Multi-step dynamics models instead condition on a sequence of future actions and predict their consequences directly, but become harder to learn as horizon grows: the model must track interacting state changes across the trajectory, endpoint supervision provides weak credit assignment, and intermediate predictions can remain plausible while losing information needed for later states. We show that these challenges can be addressed by casting the internal evolution of a multi-step transition as structured reasoning over textual world states: reasoning over sparse state changes reduces the burden of state tracking, a predictive-gain objective rewards the learned state for improving over a matched predictor that conditions on raw history instead, and intermediate predictive rewards supervise each state along the trajectory. Because these intermediate states are explicit textual representations of the world, they provide semantically meaningful targets that can be inspected, scored, and corrected during training. Across ScienceWorld, Jericho, and CEO-Bench, our approach achieves the strongest average long-horizon performance against recursive and non-recursive baselines that condition directly on raw history, with gains increasing at longer horizons. In a controlled counterfactual study, our model is also the only one with statistically significant sensitivity to future actions. --- ### Differentially Private Mixing of Public Datasets Improves Private Learning Authors: Yufei Chen, Tejumade Afonja, Anvith Thudi, Nicolas Papernot Published: 2026-10-05 Categories: cs.LG, cs.AI, cs.CR arXiv: https://arxiv.org/abs/2610.06636v1 Many machine learning applications involve sensitive data and therefore require training under differential privacy (DP). However, DP training often degrades model utility. In some cases, first pre-training the model on "public" data before finetuning with DP on the sensitive data can reduce the drop in utility. However, the success of this depends on how relevant the selected public dataset is to the sensitive data. We introduce the first pipeline that privately learns the mixture of several public datasets to pretrain on for a given sensitive downstream task. Our key insight is that we can privately find the best mixture of multiple public datasets by privately learning a low-dimensional linear model. We tested our method on the NIH dataset for X-ray classification and the ENRON email dataset for language modeling. Applying our method to find tailored mixtures of X-ray datasets to pretrain on for diseases in the NIH ChestX-ray14 dataset, we improved macro AUC by up to 0.037 across privacy budgets compared to the baselines, with gains as large as +22.8% relative AUC on Cardiomegaly at $ε=1$. For DP training on the ENRON dataset, pre-training on our mixture of The Common Pile (a collection of public-domain text datasets) decreased test perplexity by 16% relative to the baseline mixtures. --- ### Inverse Cross-spectral Neural Networks for Multivariate Time Series Authors: Lorenzo Marinucci, Leonardo Di Nino, Gabriele D'Acunto, Paolo Di Lorenzo, Sergio Barbarossa Published: 2026-10-05 Categories: stat.ML, cs.LG, eess.SP, stat.ME arXiv: https://arxiv.org/abs/2610.06630v1 CoVariance Neural Networks and their extensions have emerged as effective tools for processing multivariate data, deriving graph shift operators directly from second-order statistics. These architectures, however, are designed for independent and identically distributed observations and do not fully capture the joint structure of temporal and cross-variable dependencies in multivariate time series. In this work, we introduce Inverse Cross-Spectral Neural Networks (iCSNNs), a class of graph neural networks for stationary multivariate time series whose shift operators are the inverse cross-spectral density (iCSD) matrices. These operators encode frequency-specific conditional relationships among variables, exploiting the decomposition provided by the spectral representation theorem. Leveraging spectral smoothness, frequencies are grouped into bands sharing a single iCSD operator, yielding a compact parametrisation that retains the frequency-dependent structure of the process. We further propose a joint learning procedure to estimate both the Fourier-domain dependence structure and the iCSNN parameters, adapting the iCSD operators to the downstream task. When tested on synthetic data, iCSNN outperforms baselines from different methodological families. --- ### On the Cardinality of Optimal Representations in the Binary-Source Information Bottleneck Authors: Dier Tang, Jun Chen Published: 2026-10-05 Categories: cs.IT, cs.LG, math.OC arXiv: https://arxiv.org/abs/2610.06627v1 The information bottleneck (IB) seeks a representation $U$ of a source $X$ that retains as much information as possible about a target $Y$, subject to a constraint on $I(U;X)$. A classical argument shows that it suffices to consider representations with at most $|\mathcal{X}|+1$ symbols, and this bound is known to be tight whenever $|\mathcal{X}| \geq 3$. We show that the binary case behaves differently: if $X$ is binary and $Y$ is finite, then for every joint distribution of $(X,Y)$ and every rate constraint, the IB optimum is attained by a binary $U$. Hence the bound $|\mathcal{U}| \leq |\mathcal{X}|+1$ sharpens to $|\mathcal{U}| \leq |\mathcal{X}|$ for binary sources. The proof combines a separating hyperplane argument with the observation that, for a binary source, the ratio of the second derivatives of the two entropy functions involved is concave. ---