# Research Papers — 2026-05 · bundle page 232 of 326 > Full-text markdown of papers #11551–11600 of 16266 published in 2026-05, newest first. AI-readable mirrors of the full 16266-paper 2026-05 archive (50 papers per file). Canonical page: https://aipod.dev/papers/archive/2026-05/page/232 [← Previous bundle](https://aipod.dev/papers/archive/2026-05/page/231.md) [Next bundle →](https://aipod.dev/papers/archive/2026-05/page/233.md) ### Hyperbolic Distillation: Geometry-Guided Cross-Modal Transfer for Robust 3D Object Detection Authors: Kanglin Ning, Wenrui Li, Houde Quan, Qifan Li, Xingtao Wang, Xiaopeng Fan Published: 2026-05-11 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2605.09899v1 Cross-modal knowledge distillation has emerged as an effective strategy for integrating point cloud and image features in 3D perception tasks. However, the modality heterogeneity, spatial misalignment, and the representation crisis of multiple modalities often limit the efficient of these cross-modal distillation methods. To address these limitations in existing approaches, we propose a hyperbolic constrained cross-modal distillation method for multimodal 3D object detection (HGC-Det). The proposed HGC-Det framework includes an image branch and a point cloud branch to extract semantic features from two different modalities. The point cloud branch comprises three core components: a 2D semantic-guided voxel optimization component (SGVO), a hyperbolic geometry constrained cross-modal feature transfer component (HFT), and a feature aggregation-based geometry optimization component (FAGO). Specifically, the SGVO component adaptively refines the spatial representation of the 3D branch by leveraging semantic cues from the image branch, thereby mitigating the issue of inadequate representation fusion. The HFT component exploits the intrinsic geometric properties of hyperbolic space to alleviate semantic loss during the fusion of high-dimensional image features and low-dimensional point cloud features. Finally, the FAGO compensates for potential spatial feature degradation introduced by the 2D semantic-guided voxel optimization component. Extensive experiments on indoor datasets (SUN RGB-D, ARKitScenes) and outdoor datasets (KITTI, nuScenes) demonstrate that our method achieves a better trade-off between detection accuracy and computational cost. --- ### Pseudo-Deliberation in Language Models: When Reasoning Fails to Align Values and Actions Authors: Sushrita Rakshit, Hanwen Zhang, Hua Shen Published: 2026-05-11 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2605.09893v2 Large language models (LLMs) are often evaluated based on their stated values, yet these do not reliably translate into their actions, a discrepancy termed "value-action gap." In this work, we argue that this gap persists even under explicit reasoning, revealing a deeper failure mode we call "Pseudo-Deliberation": the appearance of principled reasoning without corresponding behavioral alignment. To study this systematically, we introduce VALDI, a framework for measuring alignment between stated values and generated dialogue. VALDI includes 4,941 human-centered scenarios across five domains, three tasks that elicit value articulation, reasoning, and action, and five metrics for quantifying value adherence. Across both proprietary and open-source LLMs, we observe consistent misalignment between expressed values and downstream dialogues. To investigate intervention strategies, we propose VIVALDI, a multi-agent value auditor that intervenes at different stages of generation. --- ### Deep Learning under Fractional-Order Differential Privacy Authors: Mohammad Partohaghighi, Roummel Marcia Published: 2026-05-11 Categories: cs.CR, cs.LG arXiv: https://arxiv.org/abs/2605.09890v1 Differentially private stochastic gradient descent (DP-SGD) is a standard approach to privacy-preserving learning based on per-example clipping, subsampling, Gaussian perturbation, and privacy accounting. Classical DP-SGD releases a noisy version of the current clipped subsampled gradient sum. We propose Fractional-Order Differentially Private Stochastic Gradient Descent (\textbf{FO-DP-SGD}), a mechanism-level extension that replaces this current-only query, before Gaussian noise is added, with a fractional recursive query combining the current clipped sum with a finite-window, power-law-weighted aggregation of previously released private sum-level outputs. This injects fractional memory into the release mechanism while preserving the standard \emph{sum-then-noise-then-divide} structure. Under add/remove adjacency with Poisson subsampling, the current-step sensitivity analysis shows that the only newly data-dependent term is the scaled current clipped sum. Hence, conditioned on the private history, the effective \(\ell_2\)-sensitivity is at most \(βC\), where \(C\) is the clipping threshold and \(β\in(0,1)\) controls the current-step contribution. Thus, FO-DP-SGD admits standard per-step Rényi differential privacy accounting via a Poisson-subsampled Gaussian mechanism with effective noise-to-sensitivity ratio \(σ/β\), and composes to yield overall \((\varepsilon,δ)\)-differential privacy guarantees. FO-DP-SGD provides a framework for studying long-memory effects in private optimization. The fractional order, memory window, and mixing coefficient govern the trade-off among current-step sensitivity, signal retention, and private-history influence. Experiments on SVHN, CIFAR-10, and CIFAR-100 show improved test accuracy and privacy--utility performance over DP-SGD and private baselines including DP-Adam, DP-IS, SA-DP-SGD, ADP-AdamW, DP-SAT, and DP-Adam-AC. --- ### The Geometric Wall: Manifold Structure Predicts Layerwise Sparse Autoencoder Scaling Laws Authors: Eslam Zaher, Maciej Trzaskowski, Quan Nguyen, Fred Roosta Published: 2026-05-11 Categories: cs.LG, cs.AI, math.DG arXiv: https://arxiv.org/abs/2605.09887v1 Sparse autoencoders (SAEs) operationalise the linear representation hypothesis: they reconstruct model activations as sparse linear combinations of interpretable dictionary atoms, on the implicit assumption that activation space is well approximated by a globally linear structure. Their reconstruction error varies sharply across layers in ways that existing scaling laws, fitted at single layers, do not explain. We argue that this variation is the empirical trace of a geometric mismatch: where the activation manifold is curved and its intrinsic dimension varies across layers, no sparse linear dictionary can match it uniformly, and the SAE's width-sparsity scaling becomes a layer-dependent function of manifold structure rather than a single universal law. We conduct the first cross-layer SAE scaling study, fitting and regressing on 844 residual-stream Gemma Scope SAE checkpoints across 68 layers of Gemma 2 2B and 9B. Stage 1 fits a per-layer scaling-law surface; Stage 2 regresses the fitted parameters and the derived per-layer width exponents on four layerwise geometric summaries. We find that manifold geometry predicts the per-layer width exponent in both models, and that the same regression coefficients learnt on one model predict the other model's per-layer exponents under cross-model transfer, indicating a transferable geometric law. At the showcase layers where richer width grids permit identification of the asymptotic floor, we find that the fitted floor tracks the layerwise geometric ordering: higher curvature and intrinsic dimension correspond to higher floor, consistent with the irreducible second-order residual that any sparse linear approximation of a curved manifold must leave behind. SAEs thus encounter not a finite-resource ceiling but a geometry-dependent wall, set by the manifold they are trying to reconstruct. --- ### The Cartesian Shortcut: Re-evaluate Vision Reasoning in Polar Coordinate Space Authors: Xia Hu, Zhenrui Yue, Brian Potetz, Howard Zhou, Leonidas Guibas, Chun-Ta Lu, Zhicheng Wang Published: 2026-05-11 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2605.09883v2 As current Multimodal Large Language Models rapidly saturate canonical visual reasoning benchmarks, a key question emerges: do these strong scores genuinely reflect robust visual understanding? We identify a pervasive vulnerability, the Cartesian Shortcut: visual reasoning benchmarks prevalently build on orthogonal grid-based layouts that can be readily discretized into explicit textual coordinates. Models systematically exploit this property, heavily leveraging text-based deductive reasoning to assist visual problem-solving. To systematically dismantle this shortcut, we introduce Polaris-Bench, which re-formulates 53 visual reasoning tasks in Polar coordinate space with paired Cartesian counterparts as reference, while preserving consistent logical constraints and task semantics -- thus fundamentally breaking the orthogonal prior that models exploit. Comprehensive evaluation across $14$ state-of-the-art MLLMs reveals that frontier models achieving $70$--$83\%$ on Cartesian layouts collapse to $31$--$39\%$ on Polar equivalents, with degradation persisting even under complete logical equivalence. Moreover, reasoning gains observed on Cartesian layouts are severely diminished on Polar equivalents. These findings expose a critical deficiency in current MLLMs: the lack of topology-invariant visual reasoning. --- ### Dissecting Jet-Tagger Through Mechanistic Interpretability Authors: Saurabh Rai, Sanmay Ganguly Published: 2026-05-11 Categories: hep-ph, cs.LG, hep-ex arXiv: https://arxiv.org/abs/2605.09881v1 Mechanistic interpretability seeks to reverse engineer a trained neural network by identifying the minimal subset of internal components. We perform a mechanistic interpretability analysis of the Particle Transformer architecture, trained on the Top Quark Tagging reference dataset, with the goal of identifying the computational circuit responsible for jet classification and characterizing the physical content of its internal representations. Combining zero ablation, path patching with two complementary on-manifold corruption strategies and linear probing of the residual stream, we identify a sparse six-head circuit that recovers the great majority of the full model performance while admitting a clean source-relay-readout interpretation. In this circuit, a single early layer head serves as the primary causal source, a cluster of middle-layer heads acts as relays selectively attending to hard pairwise substructure and a single late-layer head reads out the aggregated signal. Linear probes show that the residual stream is preferentially aligned with the energy correlator basis over the $N$-subjettiness basis. Within the energy correlator basis, the model preferentially encodes 2-prong substructure observables over the 3-prong observables. A per-layer trained probe further reveals that the apparent single step commitment of the model to a classification decision in the first class attention block is in fact a basis rotation, with the discriminating signal already saturating in the particle attention stack. These results demonstrate that mechanistic interpretability methods developed for natural language models can be used for jet physics classifiers and indicate that gradient descent may rediscover physically meaningful aspects of jet tagging without supervision. --- ### M2A: Synergizing Mathematical and Agentic Reasoning in Large Language Models Authors: Junjian Wang, Xin Zhou, Qiran Xu, Kun Zhan Published: 2026-05-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.09879v1 While reasoning has become a central capability of large language models (LLMs), the reasoning patterns required for different scenarios are often misaligned. Mathematical reasoning typically relies on intrinsic logic to solve closed-world problems in a single response, whereas agentic reasoning requires not only internal reasoning but also multi-turn interaction with external environments, interleaving thought and action. This misalignment prevents mathematical and agentic reasoning from effectively benefiting from each other, often yielding unstable reasoning behavior and only limited performance gains under multi-task learning. In this paper, we propose M2A, a novel paradigm that synergizes mathematical and agentic reasoning via model merging. To avoid overfitting to superficial reasoning patterns under joint training, M2A operates directly in parameter space: it identifies the feature subspace critical for agent behavior, and merges the mathematical reasoning task vector only along its null space, thereby injecting reasoning capability along directions that do not perturb agent behavior. Unlike SFT or RL, M2A requires no additional gradient-update and exposes the merging coefficient as a simple knob for controlling reasoning length. Experiments in a challenging real-world coding agent setting show that our method effectively extends agentic reasoning depth and delivers substantial performance improvements. Applied to a fine-tuned Qwen3-8B, M2A improves its SWE-Bench Verified resolved rate from 44.0% to 51.2% without retraining the model. Code is available at https://github.com/laplucky/M2A.git. --- ### Key-Value Means: Transformers with Expandable Block-Recurrent Compressed Memory Authors: Daniel Goldstein, Navneel Singhal, Eugene Cheah Published: 2026-05-11 Categories: cs.LG, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2605.09877v5 Recall presents a difficult choice: transformers have a linearly growing memory that slows each successive token, while linear RNNs typically have fixed costs but limited recall. We present Key-Value Means ("KVM"), a novel block-recurrence for attention that can accommodate either fixed-size or growing state. Equipping a strong transformer baseline with fixed-size KVM attention layers yields a strong $O(N)$ chunked RNN, while adding only an insignificant number of new parameters. We train a transformer with a growable KVM cache and show it performs competitively on long-context tests with only subquadratic prefill time and sublinear state growth. KVM is implementable with standard operations and without custom kernels, and supports chunk-wise parallelizable training and prefill. It provides many of the benefits of both traditional transformers (expandable context memory, chunk-wise parallelizable training and prefill) and RNNs in a single unified package. It can be used on every layer, saving KV-cache memory, and allowing a continuous range of choices of prefill time complexity between $O(N)$ and $O(N^2)$. We release our code at https://github.com/featherless-ai/KVM-paper and trained models at https://huggingface.co/collections/featherless-ai/kvm-paper under the Apache 2.0 license. --- ### Cross-Family Universality of Behavioral Axes via Anchor-Projected Representations Authors: Su-Hyeon Kim, Yo-Sub Han Published: 2026-05-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.09875v1 Large language models from different families use different hidden dimensions, tokenizers, and training procedures, making behavioral directions difficult to compare or transfer across models. We introduce an anchor-projection framework that maps hidden representations from each model into a shared anchor coordinate space (ACS). Behavioral directions extracted from source models are projected into ACS and averaged into a canonical direction. For a new model, the canonical direction is reconstructed into its native hidden space using only anchor activations, without fine-tuning or target-specific direction extraction. We evaluate five instruction-tuned model families and ten behavioral axes. We find that same-axis directions align tightly across the Llama-Qwen-Mistral-Phi (LQMP) cluster in ACS. This shared structure transfers to downstream tasks. For the aligned LQMP cluster, held-out targets achieve (0.83) ten-way detection accuracy and (0.95) mean binary AUROC, while canonical steering induces refusal-rate shifts of up to +0.46% under distribution shift. Sensitivity analyses show that two source models and small anchor pools already suffice to approximate transferable directions. Overall, ACS provides a novel perspective on cross-family interpretability, revealing that representation-level transfer remains robust across model families. --- ### EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding Authors: Ziyang Wang, Yue Zhang, Shoubin Yu, Ce Zhang, Zengqi Zhao, Jaehong Yoon, Hyunji Lee, Gedas Bertasius, Mohit Bansal Published: 2026-05-11 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2605.09874v2 Next-generation visual assistants, such as smart glasses, embodied agents, and always-on life-logging systems, must reason over an entire day or more of continuous visual experience. In ultra-long videos, relevant information is sparsely distributed across hours or days, making memory a fundamental challenge: models must accumulate information over time, recall prior states, track temporal order, and abstract recurring patterns. However, existing week-long video benchmarks are primarily designed for perception and recognition, such as moment localization or global summarization, rather than reasoning that requires integrating evidence across multiple days. To address this gap, we introduce EgoMemReason, a comprehensive benchmark for week-long egocentric video understanding through memory-driven reasoning. EgoMemReason evaluates three complementary memory types: entity memory, tracking how object states evolve and change across days; event memory, recalling and ordering activities separated by hours or days; and behavior memory, abstracting recurring patterns from sparse, repeated observations over the whole week period. EgoMemReason comprises 500 questions across three memory types and six core challenges, with an average of 5.1 video segments of evidence per question and 25.9 hours of memory backtracking. We evaluate EgoMemReason on 17 methods across MLLMs and agentic frameworks, revealing that even the best model achieves only 39.6% overall accuracy. Further analysis shows that the three memory types fail for distinct reasons and that performance degrades as evidence spans longer temporal horizons, revealing that long-horizon memory remains far from solved. We believe EgoMemReason establishes a strong foundation for evaluating and advancing long-context, memory-aware multimodal systems. --- ### EgoMemReason: A Memory-Driven Reasoning Benchmark for Long-Horizon Egocentric Video Understanding Authors: Ziyang Wang, Yue Zhang, Shoubin Yu, Ce Zhang, Zengqi Zhao, Jaehong Yoon, Hyunji Lee, Gedas Bertasius, Mohit Bansal Published: 2026-05-11 Categories: cs.CV, cs.AI, cs.CL arXiv: https://arxiv.org/abs/2605.09874v1 Next-generation visual assistants, such as smart glasses, embodied agents, and always-on life-logging systems, must reason over an entire day or more of continuous visual experience. In ultra-long video settings, relevant information is sparsely distributed across hours or days, making memory a fundamental challenge: models must accumulate information over time, recall prior states, track temporal order, and abstract recurring patterns. However, existing week-long video benchmarks are primarily designed for perception and recognition, such as moment localization or global summarization, rather than reasoning that requires integrating evidence across multiple days. To address this gap, we introduce EgoMemReason, a comprehensive benchmark that systematically evaluates week-long egocentric video understanding through memory-driven reasoning. EgoMemReason evaluates three complementary memory types: entity memory, tracking how object states evolve and change across days; event memory, recalling and ordering activities separated by hours or days; and behavior memory, abstracting recurring patterns from sparse, repeated observations over the whole week period. EgoMemReason comprises 500 questions across three memory types and six core challenges, with an average of 5.1 video segments of evidence per question and 25.9 hours of memory backtracking. We evaluate EgoMemReason on 17 methods across MLLMs and agentic frameworks, revealing that even the best model achieves only 39.6% overall accuracy. Further analysis shows that the three memory types fail for distinct reasons and that performance degrades as evidence spans longer temporal horizons, revealing that long-horizon memory remains far from solved. We believe EgoMemReason establishes a strong foundation for evaluating and advancing long-context, memory-aware multimodal systems. --- ### Intervention-Based Time Series Causal Discovery via Simulator-Generated Interventional Distributions Authors: Tsuyoshi Okita Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09870v1 We propose SVAR-FM (Structural VAR with Flow Matching), a framework for time series causal discovery that treats a physics-based simulator as a mechanical realization of Pearl's do operator. Clamping a variable inside the simulator physically severs confounding paths, producing interventional data by construction. Conditional Flow Matching then learns the nonlinear interventional conditionals. Theoretically, we prove that the full structural VAR becomes identifiable under a coverage condition on the simulator-clampable variables, and derive an end-to-end error bound that decomposes into Monte Carlo, simulator fidelity, and Flow Matching terms. A sign-flip corollary predicts that when simulator accuracy falls below a threshold, the estimated causal effect reverses sign. Empirically, a benchmark across four scientific domains confirms that SVAR-FM recovers the correct causal sign where observational methods produce sign-reversed estimates due to confounding. A case study in ultrafast laser physics verifies the sign-flip prediction by physically varying the accuracy level of a first-principles quantum solver: the low-accuracy setting reverses the causal sign, while the high-accuracy setting recovers the correct direction (R-squared = 0.983, zero bias). --- ### ConsistNav: Closing the Action Consistency Gap in Zero-Shot Object Navigation with Semantic Executive Control Authors: Haosen Wang, Zhenyang Li, Yinqiang Zhang, Zongqi He, Lutao Jiang, Kai Li, Yizhou Zhao, Liaoyuan Fan, Wenjian Hou, Tingbang Liang, Yibin Wen, Defeng Gu Published: 2026-05-11 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2605.09869v2 Zero-shot object navigation has advanced rapidly with open-vocabulary detectors, image--text models, and language-guided exploration. However, even after current methods detect a plausible target hypothesis, the agent may still oscillate between exploration and pursuit, or abandon the object near success. We identify this failure mode as an action consistency gap: semantic evidence is repeatedly reinterpreted at each step without persistent commitment across the episode. We introduce ConsistNav, a training-free zero-shot ObjectNav framework built around a semantic executive composed of three coordinated modules: Finite-State Executive Controller stages target pursuit through guarded semantic phases; Persistent Candidate Memory accumulates cross-frame target evidence into stable object hypotheses; and Stability-Aware Action Control suppresses rotational stagnation, ineffective pursuit, and unverified stopping. This design changes neither the detector nor the low-level planner; instead, it controls when semantic evidence should influence navigation and when it should be suppressed or revisited. We conduct extensive experiments on HM3D and MP3D, where ConsistNav achieves state-of-the-art results among compared zero-shot ObjectNav methods and improves SR by 11.4% and SPL by 7.9% over the controlled baseline on MP3D. Ablation studies and real-world deployment experiments further demonstrate the effectiveness and robustness of the proposed executive mechanism. --- ### Continuous Latent Contexts Enable Efficient Online Learning in Transformers Authors: Emile Anand, Abdullah Ateyeh, Xinyuan Cao, Max Dabagia Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09867v1 Large language models (LLMs) exhibit a strong capacity for in-context learning: Given labeled examples, they can generate good predictions without parameter updates. However, many interactive settings go beyond static prediction to online decision-making, in which effective behavior demands adaptation over long multi-turn horizons in response to feedback, and efficient algorithms in these domains must use compact representations of what they have learned. Recently, continuous transformer architectures with latent chain of thought have shown promise for offline iterative tasks such as directed graph-reachability. Motivated by this, we study whether continuous latent context tokens equip transformers to more effectively realize online learning. We give explicit constructions of constant-depth transformers that implement two foundational online decision-making procedures -- the weighted majority algorithm and $Q$-learning -- by storing their algorithmic state as linear combinations of feature embeddings, using a small number of latent context tokens. We further train a small GPT-2-style transformer with latent contexts using a multi-curriculum objective that does not directly supervise the latent states. On long synthetic online prediction sequences, this model outperforms larger and more complex LLMs, including Qwen-3-14B and DeepSeek-V3. Our results suggest that continuous latent contexts provide a simple and effective persistent state for transformers to implement online learning algorithms. --- ### DA-SegFormer: Damage-Aware Semantic Segmentation for Fine-Grained Disaster Assessment Authors: Kevin Zhu, William Tang, Raphael Hay Tene, Zesheng Liu, Nhut Le, Maryam Rahnemoonfar Published: 2026-05-11 Categories: cs.CV, cs.LG arXiv: https://arxiv.org/abs/2605.09864v1 Rapid and accurate damage assessment following natural disasters is critical for effective emergency response. However, identifying fine-grained damage levels (e.g., distinguishing minor from major roof damage) in UAV imagery remains challenging due to the degradation of texture cues during resizing and extreme class imbalance. We propose DA-SegFormer, a damage-aware adaptation of the SegFormer architecture optimized for high-resolution disaster imagery. Our method introduces a Class-Aware Sampling strategy to guarantee exposure to rare damage features, and it integrates Online Hard Example Mining (OHEM) with Dice Loss to dynamically focus on underrepresented classes. In addition, we employ a resolution-preserving inference protocol that maintains native texture details. Evaluated on the RescueNet dataset, DA-SegFormer achieves 74.61\% mIoU, outperforming the baseline by 2.55\%. Notably, our improvements yield double-digit gains in critical damage classes: Minor Damage (+11.7%) and Major Damage (+21.3%). --- ### Nautilus Compass: Black-box Persona Drift Detection for Production LLM Agents Authors: Chunxiao Wang Published: 2026-05-11 Categories: cs.CR, cs.AI, cs.CL, cs.IR, cs.LG arXiv: https://arxiv.org/abs/2605.09863v1 Production LLM coding agents drift over long sessions: they forget user-specified constraints, slip into mistakes the user already flagged, and confabulate prior agreements. White-box approaches such as persona vectors require model weights and so cannot be applied to closed APIs (Claude, GPT-4) that most users actually interact with. We present Nautilus Compass, a black-box persona drift detector and agent memory layer for production coding agents. The method operates entirely at the prompt-text layer: cosine similarity between user prompts and behavioral anchor texts, aggregated by a weighted top-k mean using BGE-m3 embeddings. Compass is, to our knowledge, the only public agent memory layer (among Mem0, Letta, Cognee, Zep, MemOS, smrti verified May 2026) that does not call an LLM at index time to extract facts or build a graph; raw conversation text is embedded directly. The system ships as a Claude Code plugin, an MCP 2024-11-05 A2A server (Cursor, Cline, Hermes), a CLI, and a REST API on one daemon, with a Merkle-chained audit log for tamper-evident anchor updates. On a held-out test set built from real Claude Code session traces and labeled by an independent LLM judge, Compass reaches ROC AUC 0.83 for drift detection. The embedded retrieval pipeline scores 56.6% on LongMemEval-S v0.8 and 44.4% on EverMemBench-Dynamic (n=500), topping the four published EverMemBench Table 4 baselines. LongMemEval-S 56.6% is ~30 points below recent white-box leaders (90+%); we treat that as the architectural ceiling of the no-extraction design. End-to-end reproduction cost is $3.50 (~14x cheaper than GPT-4o-judged stacks). A paired cross-vendor behavior A/B accompanies these numbers as preliminary system-level evidence. Code, anchors, frozen test data, and audit-log tooling are MIT-licensed at github.com/chunxiaoxx/nautilus-compass. --- ### UFO: A Unified Flow-Oriented Framework for Robust Continual Graph Learning Authors: Danhui Zhang, Zhe Wang, Qing Qing, Jiarui Liu, Wentao Gao, Ziqi Xu, Mingliang Hou, Xikun Zhang, Renqiang Luo Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09862v1 Graph learning research has increasingly shifted toward continual graph learning (CGL), which better reflects real-world scenarios where graphs evolve over time. However, existing CGL methods largely assume clean supervision and overlook a critical challenge: the newly arriving portions of the graph are often noisy, due to annotation errors or adversarial corruption. This mismatch limits their applicability in practice. In this work, we study robust continual graph learning, where models must simultaneously handle catastrophic forgetting and noisy supervision in evolving graph data. We show that label noise introduces a new failure mode, catastrophic remembering, where models persistently reinforce corrupted knowledge across tasks. To address these challenges, we propose a Unified Flow-Oriented framework (UFO). First, UFO models conditional feature distributions via flow-based generative modeling and produces replay representations, mitigating forgetting without storing historical data. Second, UFO estimates instance-level reliability scores to distinguish clean from noisy nodes, reducing the impact of corrupted supervision and alleviating catastrophic remembering. Extensive experiments on four benchmark graph datasets under varying noise ratios demonstrate that UFO consistently outperforms existing methods in both accuracy and forgetting metrics. Code is available at: https://anonymous.4open.science/r/UFO. --- ### Flag Varieties: A Geometric Framework for Deep Network Alignment Authors: Jingchuan Xiao, Xinyi Sui, Cihan Ruan Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09861v1 Alignment, the tendency of adjacent weight matrices in deep networks to develop compatible subspace orientations, underlies gradient flow, Neural Collapse, and representation similarity across architectures. Despite extensive empirical documentation, these phenomena have resisted unified theoretical treatment: existing explanations are post-hoc, each fitted to a specific observation with whatever mathematics is at hand. We reverse this direction by deriving the mathematical structure that layerwise alignment inherently demands. Using geometric invariant theory, we prove that alignment geometry has a canonical closed, polystable stratum given by a flag variety, and that subspace intersection dimension is its unique reparameterization-invariant observable, establishing that subspace metrics are not empirical conventions but mathematical necessities. This unified framework yields two dynamical consequences: ridge regularization drives subspace alignment at an exponential rate set by weight decay, whereas nonlinear activations induce a commutator obstruction to exact basis alignment, generically present in nonlinear networks and absent in linear ones. Together these give a geometric explanation of the Level-2/3 hierarchy in Neural Collapse from first principles rather than post-hoc analysis. The commutator magnitude and head subspace overlap further serve as weight-space windows into internal alignment structure, requiring no forward passes. Experiments on multilayer perceptrons, residual networks, and pretrained language models support the proposed diagnostics and delineate their scope. --- ### PACE: Policy-Native Adaptive Decision Timing for Long-Horizon Reasoning Authors: Chen Li, Zhantao Yang, Fangyi Chen, Han Zhang, Anudeepsekhar Bolimera, Marios Savvides Published: 2026-05-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.09860v4 Long-horizon reasoning requires deciding not only what actions to take, but how many to execute open-loop before replanning. This number, the execution depth, balances replanning cost against compounding execution errors. Most current systems either fix the execution depth as a hand-tuned scalar or adjust it at inference time with heuristic rules decoupled from the policy; we argue both can be suboptimal. In this work, we treat the execution depth as a learnable, history-conditioned variable of the policy itself, and propose PACE, a model-native VLM policy that jointly predicts what to execute and for how long under a hard decision budget. We evaluate PACE on four long-horizon environments spanning fully and partially observable settings and visual and textual modalities: Sliding Puzzle, Sokoban, ALFWorld, and ScienceWorld. Across all settings, PACE improves success rate by 3.1 to 15.6 percentage points, while using fewer decisions on average, achieving empirical success-decision Pareto dominance. --- ### When to Re-Commit: Temporal Abstraction Discovery for Long-Horizon Vision-Language Reasoning Authors: Chen Li, Zhantao Yang, Fangyi Chen, Han Zhang, Anudeepsekhar Bolimera, Marios Savvides Published: 2026-05-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.09860v3 Long-horizon reasoning requires deciding not only what actions to take, but how deeply to commit before the next observation. We formalize this as \emph{commitment depth}: the number of primitive actions executed open-loop between replans. Commitment depth induces a trade-off between replanning cost and compounding execution error, yet most existing long-horizon systems fix it as a hand-designed scalar. In this work, we instead treat commitment depth as a learnable, state-conditioned variable of the policy itself. We instantiate this within a model-native vision--language policy that jointly predicts both what to execute and for how long. Across Sliding Puzzle and Sokoban, the resulting adaptive policy Pareto-dominates every non-degenerate fixed-depth baseline, achieving up to 12.5 percentage points higher solve rate while using approximately 25\% fewer primitive actions per episode. Despite using a 7B backbone, our method outperforms GPT-5.5 and Claude Sonnet on both tasks, while every tested open-weight vision--language model achieves 0\% zero-shot success. We further present a theoretical analysis showing that, under the standard commitment-depth surrogate, state-conditioned commitment strictly dominates any fixed depth whenever the locally optimal depth varies across states. --- ### Learning to Align Generative Appearance Priors for Fine-grained Image Retrieval Authors: Shijie Wang, Yadan Luo, Zijian Wang, Xin Yu, Zi Huang Published: 2026-05-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.09859v1 Fine-grained image retrieval (FGIR) typically relies on supervision from seen categories to learn discriminative embeddings for retrieving unseen categories. However, such supervision often biases retrieval models toward the semantics of seen categories rather than the underlying appearance characteristics that generalize across categories, thereby limiting retrieval performance on unseen categories. To tackle this, we propose GAPan, a Generative Appearance Prior alignment network that reformulates the learning objective from category prediction toward appearance modeling. Technically, GAPan treats retrieval features with an invertible density model based on normalizing flows. In the forward direction, the flow maps all instance features into a latent density space, where each seen category is modeled by a class-conditional Gaussian prior and optimized via exact likelihood estimation. This formulation preserves richer appearance details by leveraging the invertible property of the flows. In the reverse direction, samples from the high-density regions of these learned priors are mapped back to the feature space to produce appearance-aware anchors that reflect intra-category variation. These anchors supervise a prior-driven alignment objective that aligns retrieval embeddings with category-specific appearance distributions, thereby improving generalization to unseen categories. Evaluations demonstrate that our GAPan achieves state-of-the-art performance on both widely-used fine- and coarse-grained benchmarks. --- ### Clip-level Uncertainty and Temporal-aware Active Learning for End-to-End Multi-Object Tracking Authors: Riku Inoue, Shogo Sato, Kazuhiko Murasaki, Tomoyasu Shimada, Toshihiko Nishimura, Ryuichi Tanida Published: 2026-05-11 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.09858v1 Multi-Object Tracking (MOT) in dynamic environments relies on robust temporal reasoning to maintain consistent object identities over time. Transformer-based end-to-end MOT models achieve strong performance by explicitly modeling temporal dependencies, yet training them requires extensive bounding-box and identity annotations. Given the high labeling cost and strong redundancy in videos, Active Learning (AL) is an effective approach to improve annotation efficiency. However, existing AL methods for MOT primarily operate at the frame level, which is structurally misaligned with modern end-to-end trackers whose inference and training rely on multi-frame clips. To bridge this gap, we formulate clip-level active learning and propose Clip-level Uncertainty and Temporal-aware Active Learning (CUTAL). In contrast to frame-based approaches, CUTAL scores each clip using uncertainty metrics derived from multi-frame predictions to capture inter-frame correspondence ambiguities, while enforcing temporal diversity to select an informative and non-redundant subset. Experiments show that CUTAL achieves stronger overall performance than baselines at the same label budgets across MeMOTR and SambaMOTR. Notably, CUTAL achieves performance comparable to full supervision for MeMOTR on both datasets using only 50% of the labeled training data. --- ### Unified Approach for Weakly Supervised Multicalibration Authors: Futoshi Futami, Takashi Ishida Published: 2026-05-11 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2605.09857v1 Multicalibration requires predicted scores to agree with label probabilities across rich families of subgroups and score-dependent tests, but existing methods require clean input-label pairs for evaluation and post-processing. This assumption fails in weakly supervised learning (WSL) regimes -- including positive-unlabeled, unlabeled-unlabeled, and positive-confidence learning -- where clean labels are costly or unavailable even though reliable uncertainty estimates may be crucial. We address this gap by developing estimators of multicalibration error and post-hoc correction methods for WSL settings in which clean input-label pairs are unavailable. We propose a unified framework for estimating and correcting multicalibration under weak supervision by combining contamination-matrix risk rewrites with witness-based calibration constraints, yielding corrected multicalibration moments with finite-sample guarantees. We further propose weak-label multicalibration boost (WLMC), a generic post-hoc recalibration algorithm under weak supervision. Finally, we conduct experiments across multiple weak-supervision settings to evaluate multicalibration behavior and offer empirical insight into uncertainty estimation under weak supervision. --- ### MoPO: Incorporating Motion Prior for Occluded Human Mesh Recovery Authors: Tao Tang, Hong Liu, Xinshun Wang, Wanruo Zhang Published: 2026-05-11 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2605.09856v1 Although recent studies have made remarkable progress in human mesh recovery, they still exhibit limited robustness to occlusions and often produce inaccurate poses and severe motion jitter due to the insufficient spatial features for occluded body parts. Inspired by the rapid advancements in human motion prediction, we discover that compared to occluded image features, pose sequence inherently contains reliable motion prior for estimating occluded body parts. In this paper, we incorporate Motion Prior for Occluded human mesh recovery, called MoPO. Our MoPO mainly consists of two components: 1) The motion de-occlusion module, where we propose a spatial-temporal occlusion detector to detect joint visibility, and then we propose a lightweight motion predictor to complete the occluded body parts by predicting the most plausible joint positions based on history poses. 2) The motion-aware fusion and refinement module, which fuses the completed joint sequence with image features to estimate human shape and initial human pose. Moreover, the completed joint sequence is further used to refine the final human pose through inverse kinematics, which provides the occlusion-free motion prior for regressing human poses. Extensive experiments demonstrate that MoPO achieves state-of-the-art performance on both occlusion-specific and standard benchmarks, significantly enhancing the accuracy and temporal consistency of occluded human mesh recovery. Our code and demo can be found in the supplementary material. --- ### Concordia: Self-Improving Synthetic Tables for Federated LLMs Authors: Jimin Huang, Duanyu Feng, Nuo Chen, Xiaoyu Wang, Zhiqiang Zhang, Xueqing Peng, Mingquan Lin, Prayag Tiwari, Guojun Xiong, Alejandro Lopez-Lira, Sophia Ananiadou Published: 2026-05-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.09855v2 Federated learning (FL) enables training large language models (LLMs) without sharing raw data, but adapting LLMs under strict data isolation and non-IID client distributions remains challenging in practice. Synthetic data offers a natural privacy-preserving surrogate for local training, yet existing federated pipelines typically treat synthetic generation as static or loosely coupled with downstream optimization, leading to rapidly diminishing utility under heterogeneous clients. We study federated adaptation of LLMs on tabular tasks where raw records and validation data cannot be shared, and local training must rely entirely on synthetic tables. We propose Concordia, a tri-level optimization framework that aligns synthetic data generation with federated validation utility despite these constraints. At the client level, models are adapted via parameter-efficient LoRA training on synthetic tables. Clients additionally learn lightweight utility scorers from private validation feedback to reweight synthetic samples during local training. At the outer level, each client refines its own synthetic table generator using group-relative policy optimization (GRPO), guided by an ensemble of heterogeneous scorers shared across clients, without aggregating generator parameters or exposing validation data. Experiments on privacy-sensitive tabular benchmarks from finance and healthcare demonstrate that Concordia consistently improves federated performance, cross-client stability, and robustness to distribution shift compared to static and decoupled synthetic-data baselines. --- ### Exploration-Driven Optimization for Test-Time Large Language Model Reasoning Authors: Changhao Li, Yuchen Zhuang, Chenxiao Gao, Haotian Sun, Rushi Qiang, Chao Zhang, Bo Dai Published: 2026-05-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.09853v1 Post-training techniques combined with inference-time scaling significantly enhance the reasoning and alignment capabilities of large language models (LLMs). However, a fundamental tension arises: inference-time methods benefit from diverse sampling from a relatively flattened probability distribution, whereas reinforcement learning (RL)-based post-training inherently sharpens these distributions. To address this, we propose Exploration-Driven Optimization (EDO), which extends reward-biasing style exploration objectives to iterative post-training and integrates them into standard RL objectives, encouraging greater diversity in sampled solutions while facilitating more effective inference-time computation. We incorporate EDO into iterative Direct Preference Optimization (iDPO) and Group Relative Policy Optimization (GRPO), resulting in two variants: ED-iDPO and ED-GRPO. Extensive experiments demonstrate that both ED-iDPO and ED-GRPO exhibit greater solution diversity and improved reasoning abilities, particularly when combined with test-time computation techniques like self-consistency. Across three in-distribution reasoning benchmarks, EDO achieves a 1.0-1.3\% improvement over the strongest baselines, and delivers an additional 1.5\% average gain on five out-of-distribution tasks. Beyond accuracy, EDO preserves model entropy and stabilizes RL training dynamics, highlighting its effectiveness in preventing over-optimization collapse. Taken together, these results establish EDO as a practical framework for balancing exploration and exploitation in LLM reasoning, especially in settings that rely on test-time scaling. --- ### Fairness of Explanations in Artificial Intelligence (AI): A Unifying Framework, Axioms, and Future Direction toward Responsible AI Authors: Gideon Popoola, John Sheppard Published: 2026-05-11 Categories: cs.AI, cs.CE, cs.CY, cs.LG arXiv: https://arxiv.org/abs/2605.09852v1 Machine learning algorithms are being used in high-stakes decisions, including those in criminal justice, healthcare, credit, and employment. The research community has responded with two largely independent research fields: \emph{algorithmic fairness}, which targets equitable outcomes, and \emph{explainable AI} (XAI), which targets interpretable reasoning. This survey identifies and maps a novel blind spot at their intersection, which is a model that can satisfy every standard fairness criterion in its outputs while being profoundly unfair in its \emph{reasoning process}. We refer to this as the procedural bias, and mitigating it requires treating the fairness of explanations as a distinct object of scientific study. To our knowledge, we provide the first unified theoretical and literature review of this emerging field and elucidate the drawbacks of post-hoc explainers in certifying explanation fairness. Our central contribution is a \emph{conditional invariance framework} formalizing explanation fairness as the requirement that explanations should be indifferent regardless of the protected attributes $ P(E(X) \in \cdot \mid X_\text{rel} = x_\text{rel},\, A = a) = P(E(X) \in \cdot \mid X_\text{rel} = x_\text{rel},\, A = b)$ for all task-relevant $x$, a single principle from which all existing explanation fairness metrics emerge as partial operationalizations. We introduce a seven-dimensional taxonomy, identify three generative mechanisms of explanation inequity (representation-driven, explanation-model mismatch, actionability-driven), and propose a canonical six-step evaluation workflow for operationalizing explanation fairness audits in practice. --- ### Probing Routing-Conditional Calibration in Attention-Residual Transformers Authors: Wenhao Liang, Lin Yue, Wei Emma Zhang, Miao Xu, Mingyu Guo, Olaf Maennel, Weitong Chen Published: 2026-05-11 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2605.09850v1 Post-hoc calibration is usually evaluated as a function of logits or softmax confidence alone, even as routing-augmented architectures increasingly accompany predictions with sample-specific internal routing traces and pair them with claims of calibration-relevant uncertainty. We ask a basic question: do these traces provide stable routing-specific evidence for post-hoc calibration beyond confidence? We study this in Attention-Residual transformers (Kimi Team, 2026) through a matched-confidence diagnostic suite that stratifies examples by routing-derived state, compares subgroup gaps against within-bin routing-permutation nulls, and evaluates matched post-hoc probes differing only in their auxiliary feature. Across our completed AR runs, scalar routing summaries do not provide stable evidence of routing-conditional miscalibration: weighted gaps remain small or seed-sensitive, and only $1$ of $30$ within-bin permutation tests rejects the conditional-null at $α=0.05$ (only on one seed; not stable across seeds in that cell). AR-CondCal, a minimal $2$-D Nadaraya--Watson probe on confidence and routing-depth variance, lies within the seed-variance band of matched confidence-only and predictive-entropy controls and does not reliably improve worst-routing-tertile ECE; bandwidth-sensitivity checks (Scott multiples, CV-NLL, global-ECE oracle) do not change this. A full-vector MLP over $(c, H_1, \ldots, H_L)$ can appear to improve over a linear confidence baseline, but the apparent gain disappears once a capacity-matched confidence-only MLP is included as a control, and shuffled routing profiles achieve comparable performance. Apparent routing-aware calibration gains in this AR setting should not be read as internal-state calibration until matched-confidence, bandwidth, capacity, and permutation controls rule out common confounds. --- ### Efficient Neural Architectures for Real-Time ECG Interpretation on Limited Hardware Authors: Ashery Mbilinyi, Callum O'Riley, Julia Handra, Ashley Moller-Hansen, Jason Andrade, Marc Deyell, Cameron Hague, Nathaniel Hawkins, Kendall Ho, Jonathan Leipsic, Roger Tam Published: 2026-05-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.09848v1 Electrocardiogram (ECG) interpretation is essential for diagnosing a wide range of cardiac abnormalities. While deep learning has shown strong potential for automating ECG classification, many existing models rely on large, computationally intensive architectures that hinder practical deployment. In this paper, we present an empirical study of convolutional neural network (CNN) architectures, exploring tradeoffs between diagnostic accuracy and computational efficiency. We benchmark two established baselines: AttiaNet, a compact model composed of sequential temporal and spatial blocks, and DeepResidualCNN, the winning architecture of the 2021 PhysioNet/Computing in Cardiology Challenge. Building on these, we propose three lightweight models: (i) ParallelCNN, which employs dual temporal and spatial branches for parallel pattern extraction; (ii) ParallelCNNew, a variant with symmetric weight initialization for balanced feature learning; and (iii) SimpleNet, a streamlined architecture that jointly processes temporal and spatial dimensions. Our experiments span three publicly available 12-lead ECG datasets from Germany, China, and the United States, covering binary, multiclass, and multilabel classification tasks across diverse patient populations. We further evaluate the impact of integrating low-cost demographic metadata (age and sex) to improve performance with minimal overhead. To ensure fair comparison, we introduce a unified Efficiency Score that integrates model size, inference speed, memory usage, and AUC performance. By balancing diagnostic performance and efficiency, our models offer a scalable and viable foundation for next-generation AI systems in cardiovascular care. --- ### ChladniSonify: A Visual-Acoustic Mapping Method for Chladni Patterns in New Media Art Creation Authors: Yakun Liu, Hai Luan, Dong Liu, Zhiyu Jin Published: 2026-05-11 Categories: cs.SD, cs.AI arXiv: https://arxiv.org/abs/2605.09846v1 In new media art creation, the mapping between vision and hearing is often subjective. As a classic carrier of sound visualization, Chladni patterns have great potential in building audio-visual mapping mechanisms. However, existing tools face pain points: high technical barriers for simulation, offline computing failing real-time interaction, and uncontrollable mapping rules in general sonification tools. To address these, this paper proposes ChladniSonify, a real-time visual-acoustic mapping method for Chladni patterns. Based on Kirchhoff-Love plate theory, we build a paired dataset via numerical programming and calibrate it using ANSYS finite element simulation. Focusing on the slender nodal lines of Chladni patterns, we adopt a lightweight CNN with CBAM to achieve high-precision, low-latency pattern classification. Finally, we build an end-to-end system in Python and Max/MSP, mapping recognized patterns to corresponding sine wave frequencies. Results show the system has excellent usability: the classification module achieves 99.33% accuracy on the test set with 7.03 ms inference latency; the mapped frequency matches the theoretical value with zero deviation; the average end-to-end latency is under 50 ms, meeting real-time interactive needs. This work provides a reproducible engineering prototype for Chladni audio-visual art creation. --- ### Sub-Footprint Effect Correction in FW-LiDAR Point Clouds via Intra-Footprint Target Unmixing Authors: Zhen Xiao, Yanfeng Gu, Xian Li Published: 2026-05-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.09845v1 Sub-footprint target mixing within a laser footprint significantly increases LiDAR intensity uncertainty, especially in complex environments where heterogeneous materials inside one footprint cause nonlinear distortions that impair intensity-based applications. However, the forward mixing inherent to the single-pixel detection mode of LiDAR systems blurs sub-footprint contributions, making sub-footprint effects difficult to address effectively in existing studies. To address this issue, we introduce a novel, physics-based framework that explicitly resolves sub-footprint intensity correction in full-waveform LiDAR (FW-LiDAR) point clouds. The key innovation is to make the otherwise implicit intra-footprint mixing process explicit: we first develop a spatiotemporal laser-beam distribution model to physically characterize within-footprint forward mixing of multi-target returns. Building on this formulation, we incorporate ancillary information including waveform parameters and surface geometry as constraints to pose a well-defined inverse unmixing problem and decompose each footprint into fractional contributions from multiple sub-targets. We then recover sub-footprint-corrected intensities by inverting the observed mixtures through a unified combination of parametric and model-driven approaches. To the best of our knowledge, few prior studies explicitly establish sub-footprint inversion and correction within a single laser footprint, and our framework offers a principled, physics-grounded solution. Experiments on both controlled and real-world LiDAR datasets demonstrate that the proposed method significantly enhances semantic separability across heterogeneous targets and intensity consistency across homogeneous targets. --- ### The Metacognitive Probe: Five Behavioural Calibration Diagnostics for LLMs Authors: Rafael C. T. Oliveira Published: 2026-05-11 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2605.09844v1 The Metacognitive Probe is an exploratory five-task, 15-slot diagnostic that decomposes an LLM's confidence behaviour into five behaviourally-distinct dimensions: confidence calibration (T1-CC), epistemic vigilance (T2-EV), knowledge boundary (T3-KB), calibration range (T4-CR), and reasoning-chain validation (T5-RCV). It is evaluated on N=8 frontier models and N=69 humans. The instrument is motivated by Flavell (1979) and Nelson and Narens (1990) but operates on observable confidence-correctness alignment; it is not a validated cross-species metacognition scale, and the pre-specified human developmental hypothesis was falsified. Composite benchmarks (MMLU, BIG-Bench, HELM, GPQA) ask whether a model produces a correct response. They are silent on whether the model knows when its response is wrong. A model can score 80 on a composite calibration benchmark and still be wildly overconfident in narrow pockets the aggregate cannot surface. The Metacognitive Probe surfaces those pockets. Our headline is a 47-point within-model dissociation in Gemini 2.5 Flash: panel-best within-task calibration (T1-CC = 88; Spearman rho = +0.551, 95% CI (+0.14, +0.80), p = 0.005) and panel-worst cross-task difficulty prediction (T4-CR = 41; sigma_conf = 1.4 across twelve factoids). --- ### Yield Curve Forecasting using Machine Learning and Econometrics: A Comparative Analysis Authors: Aman Singh, Tokunbo Ogunfunmi, Sanjiv Das Published: 2026-05-11 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.09842v1 While machine learning has revolutionized many fields such as natural language processing (NLP) and computer vision, its impact on time-series forecasting is still widely disputed, especially in the finance domain. This paper compares forecasting performance on U.S. Treasury yield curve data across econometrics/time-series analysis, classical machine learning, and deep learning methods, using daily data over 47 years. The Treasury yield curve is important because it is widely used by every participant in the bond markets, which are larger than equity markets. We examine a variety of methods that have not been tested on yield curve forecasting, especially deep learning algorithms. The algorithms include the Autoregressive Integrated Moving Average (ARIMA) model and its extensions, naive benchmarks, ensemble methods, Recurrent Neural Networks (RNNs), and multiple transformers built for forecasting. ARIMA and naive econometric models outperform other models overall, except in one time block. Of the machine learning methods, TimeGPT, LGBM and RNNs perform the best. Furthermore, the paper explores whether stationary or nonstationary data are more appropriate as input to deep learning models. --- ### Free Energy Manifold: Score-Based Inference for Hybrid Bayesian Networks Authors: Cheol Young Park, Shou Matsumoto Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09839v1 We introduce the Free Energy Manifold (FEM), a score-trained conditional energy model specialized for inference in hybrid Bayesian networks with discrete and continuous variables. FEM represents each conditional factor as an energy landscape over learned discrete-parent embeddings and continuous observations, enabling posterior evaluation, generative sampling, and compositional inference across multiple continuous leaves by energy addition under conditional independence. A central finding is the mode-bridge artifact: standard conditional energy models can create low-energy ridges between separated modes of the same class, producing overconfident posteriors at off-data interior points. We analyze this failure and propose valley regularization, an off-data calibration term that restores near-uniform posteriors in such regions while preserving in-data fit. Across synthetic multimodal hybrid-BN benchmarks, FEM substantially reduces KL divergence relative to classical baselines and a vanilla conditional EBM, including large gains at mode-bridge midpoint queries and in multi-leaf evidence composition. We also evaluate high-cardinality discrete-parent settings and a UCI Breast Cancer sanity check, showing that FEM is most useful when multimodal or compositional Bayesian-network inference is required, while discriminative classifiers remain preferable for closed-world classification tasks. --- ### The Association of Transformer-based Sentiment Analysis with Symptom Distress and Deterioration in Routine Psychotherapy Care Authors: Douglas K. Faust, Peter Awad, Alexandre Vaz, Tony Rousmaniere Published: 2026-05-11 Categories: cs.CL, cs.LG arXiv: https://arxiv.org/abs/2605.09838v1 Sentiment analysis has been of long-standing interest in psychotherapy research. Recently, the Transformer deep learning architecture has produced text-based sentiment analysis models that are highly accurate and context-aware. These models have been explored as proxies for emotion measurement instruments in psychotherapy, but not investigated as stand-alone psychometric tools. Using proposed utterance-level and session-level sentiment features derived from a fine-grained sentiment model on a large corpus of psychotherapy sessions (N = 751), we investigate the distribution of session aggregated sentiment scores. Further, we characterize the relationship of these features to individual components and the overall score of the OQ-45 instrument and find that this sentiment feature is most strongly correlated to components related to emotional valence in directionally intuitive ways. Finally, we report that there are statistically significant differences between the sentiment distributions for patients flagged as at risk of deterioration or dropping out of care via either the OQ Rational or Empirical outcome models. These correlations to a fully-validated psychometric instrument demonstrate that these proposed sentiment features are, at least, adjunctive measures of client distress and deterioration. --- ### Supercharging Bayesian Inference with Reliable AI-Informed Priors Authors: Jongwoo Choi, Sean O'Hagan Published: 2026-05-11 Categories: stat.ML, cs.LG arXiv: https://arxiv.org/abs/2605.09834v1 Modern predictive systems encode beliefs that can act as useful prior information for statistical inference in data-limited settings. Using them for prior construction introduces a tradeoff: an informative prior built from a predictive model can sharpen inference from limited data, but also risks propagating error from the model into the posterior. We propose a framework for AI-informed prior elicitation that mitigates this tension by rectifying the AI-induced law that generates synthetic data before using it to inform a prior. The rectified law can be embedded into synthetic data-driven prior elicitation techniques, including as a base measure in a Dirichlet process (DP) prior on the data-generating process. We refer to the resulting prior and corresponding posterior as the rectified AI prior and rectified AI posterior. We establish Gaussian asymptotics for the rectified AI posterior under non-vanishing prior strength and derive a first-order expression for its centering bias. Our rectified AI priors substantially reduce bias compared to standard approaches, improve the coverage of credible intervals, and make AI-powered prior information more reliable. We additionally apply the rectified AI prior to a real skin disease classification task and show that it can meaningfully boost predictive performance. --- ### Cross-Domain Lossy Compression via Constrained Minimum Entropy Coupling Authors: Nam Nguyen, Hassan Tavakoli, An Vuong, Thinh Nguyen, Bella Bose Published: 2026-05-11 Categories: cs.IT, cs.LG arXiv: https://arxiv.org/abs/2605.09833v1 This paper studies cross-domain lossy compression through the lens of minimum entropy coupling (MEC) with rate and classification constraints. In this setting, an encoder observes samples from a degraded source domain, while the decoder is required to generate outputs following a prescribed target distribution and to preserve information relevant to a downstream classification task. Motivated by logarithmic-loss distortion, we adopt an information-based objective that maximizes the coupling strength between the source and reconstruction, rather than minimizing a sample-wise distortion. Under common randomness, we formulate a rate-constrained MEC problem (MEC-B) and show that the intermediate representation can be removed without loss of optimality, yielding an equivalent deterministic coupling formulation. For Bernoulli sources, closed-form expressions are derived with and without classification constraints. In addition, we implement a neural restoration framework using quantization, entropy modeling, distribution matching, and classification regularization. Experiments on MNIST super-resolution and SVHN denoising show that increasing the available rate improves classification accuracy and yields more informative reconstructions. --- ### Modeling Atomic Conformational Ensembles of Proteins via Test-Time Supervision of Boltz-2 on Cryo-EM Density Maps Authors: Jay Shenoy, Miro Astore, Axel Levy, Frédéric Poitevin, Sonya M. Hanson, Gordon Wetzstein Published: 2026-05-11 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.09832v1 Knowledge of a protein's atomic conformational ensemble is critical to determining its function, yet state-of-the-art ensemble prediction models are limited by lack of high-quality conformational data from simulation or experiment. Recent advances in heterogeneous reconstruction for cryo-electron microscopy (cryo-EM) have enabled scientists to visualize ensembles of density maps for larger proteins and complexes not typically accessible through simulation, but building atomic models into these maps remains a challenge. Traditionally, ensemble prediction models are trained via a two-stage process: experimental density maps are converted into atomic structural ensembles through model building, after which these structures are used to train sequence-to-atomic ensemble predictors. In this work, we propose a new principle for fine-tuning pre-trained static structure prediction models such as Boltz-2 directly on raw cryo-EM maps, bypassing the two-stage process. We apply this technique to the problem of atomic model building by fine-tuning Boltz-2 to generate atomic conformations from an input ensemble of cryo-EM maps, achieving superior model building accuracy compared to prior work. Beyond overfitting to individual map ensembles, our method, CryoSampler, also shows preliminary evidence of in-domain generalization after fine-tuning, sampling diverse atomic conformations for an unseen sequences within the same protein family without requiring cryo-EM data. These capabilities indicate that CryoSampler holds the potential to train next-generation atomic ensemble prediction models directly on raw cryo-EM measurements. --- ### Loom: Hybrid Retrieval-Scoring Outfit Recommendation with Semantic Material Compatibility and Occasion-Aware Embedding Priors Authors: Anushree Berlia Published: 2026-05-11 Categories: cs.IR, cs.CV arXiv: https://arxiv.org/abs/2605.09830v1 We present Loom, an outfit recommendation system that combines neural embedding retrieval with structured domain scoring to generate complete, coherent outfits from fashion catalogs. Given an anchor clothing item, Loom retrieves complementary pieces via slot-constrained approximate nearest neighbor search over FashionCLIP embeddings, then scores candidate outfits using a multi-objective function that integrates six signals: embedding similarity, color harmony, formality consistency, occasion coherence, style direction, and within-outfit diversity. We introduce two techniques that address limitations of purely learned or purely rule-based approaches: (1) semantic material weight, which uses CLIP embedding geometry to infer garment heaviness for layer compatibility without hand-coded material taxonomies; and (2) vibe/anti-vibe occasion priors, which embed prose descriptions of occasion contexts as anchor vectors in CLIP space and score items by differential affinity. Ablation experiments on a catalog of 620 items show that each component contributes measurably to outfit quality: the full system achieves a mean outfit score of 0.179 with a 9.3% hard violation rate, compared to 0.054 score and 16.0% violations for a category-constrained random baseline, a 3.3x improvement in score and 42% reduction in violations. Direction reranking is the single indispensable component: removing it drops score to 0.052, essentially equal to random. The system generates three stylistically distinct outfits in under 5 seconds on commodity hardware. --- ### Fashion Florence: Fine-Tuning Florence-2 for Structured Fashion Attribute Extraction Authors: Anushree Berlia Published: 2026-05-11 Categories: cs.CV, cs.AI arXiv: https://arxiv.org/abs/2605.09827v1 We present Fashion Florence, a Florence-2 vision-language model fine-tuned with LoRA to extract structured fashion attributes from clothing images. Given a single photograph, the model generates a JSON object containing category, color, material, style tags, and occasion tags, structured output suitable for direct programmatic consumption by downstream recommendation and retrieval systems. Fine-tuning data is derived from the iMaterialist Fashion dataset (228 labels), where we collapse fine-grained annotations into a compact 6-category, 16-color, 19-style schema via rule-based label engineering. We apply LoRA (r=16, alpha=32) to all decoder linear layers, training for 3 epochs on 3,688 examples. On a held-out test set of 461 images, Fashion Florence achieves 94.6% category accuracy and 63.0% material accuracy, compared to 89.3% / 43.3% for GPT-4o-mini and 87.4% for Gemini 2.5 Flash. Fashion Florence produces valid JSON in 99.8% of outputs while running at 0.77B parameters on a single GPU at zero marginal inference cost. Style tag F1 reaches 0.753 vs. 0.612 (Gemini) and 0.398 (GPT-4o-mini). The model is deployed as a Hugging Face Space and integrated into Loom, an open-source outfit recommendation system. --- ### EnactToM: An Evolving Benchmark for Functional Theory of Mind in Embodied Agents Authors: Gurusha Juneja, Dylan Lu, Saaket Agashe, Parth Diwane, Edward Gunn, Jayanth Srinivasa, Gaowen Liu, William Yang Wang, Yali Du, Xin Eric Wang Published: 2026-05-11 Categories: cs.AI, cs.MA arXiv: https://arxiv.org/abs/2605.09826v2 Theory of Mind (ToM), the ability to track others epistemic state, makes humans efficient collaborators. AI agents need the same capacity in multi agent settings, yet existing benchmarks mostly test literal ToM by asking direct belief questions. The ability act optimally on implicit beliefs in embodied environments, called functional ToM, remains largely untested. We introduce EnactToM, an evolving benchmark of 300 embodied multi-agent tasks set in a 3D household with partial observability, private information, and constrained communication. Each task is formally verified for solvability and required epistemic depth, and new tasks are generated increase difficulty as models improve. On the hard split, all seven evaluated frontier models score 0.0% Pass^3 on functional task completion, while averaging 45.0% on literal belief probes. Manual analysis traces 93% of sampled failures to epistemic coordination breakdowns such as withheld information, ignored partner constraints, and misallocated messages, providing a concrete target for future work. --- ### Pretraining large language models with MXFP4 on Native FP4 Hardware Authors: Musa Cim, Sarthak Arora, Poovaiah Palangappa, Miro Hodak, Ravi Dwivedula, Meena Arunachalam, Mahmut Taylan Kandemir Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09825v4 Why does full-pipeline FP4 training of large language models often diverge, even when forward activations and activation gradients remain stable? We address this question through a controlled study of MXFP4 quantization in transformer training, progressively enabling FP4 across forward propagation (Fprop), activation gradients (Dgrad), and weight gradients (Wgrad) while holding all other factors fixed. In full pretraining of Llama 3.1-8B on the C4 dataset, we observe that quantizing Wgrad is the primary driver of convergence degradation, whereas FP4 in Fprop and Dgrad alone introduces only modest additional token requirements. To interpret this behavior, we evaluate both structured and stochastic interventions under a controlled experimental setting. We find that stochastic rounding and randomized Hadamard rotations fail to stabilize training once Wgrad is quantized, whereas deterministic Hadamard rotations consistently restore stable optimization. These results suggest that FP4 training instability is driven by structured micro-scaling errors along sensitive gradient paths, rather than by insufficient stochasticity. We run experiments with native MXFP4 support on AMD Instinct MI355X GPUs, enabling controlled investigation of these effects without reliance on software emulation. --- ### Pretraining large language models with MXFP4 on Native FP4 Hardware Authors: Musa Cim, Poovaiah Palangappa, Miro Hodak, Ravi Dwivedula, Meena Arunachalam, Mahmut Taylan Kandemir Published: 2026-05-11 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.09825v3 Why does full-pipeline FP4 training of large language models often diverge, even when forward activations and activation gradients remain stable? We address this question through a controlled study of MXFP4 quantization in transformer training, progressively enabling FP4 across forward propagation (Fprop), activation gradients (Dgrad), and weight gradients (Wgrad) while holding all other factors fixed. In full pretraining of Llama 3.1-8B on the C4 dataset, we observe that quantizing Wgrad is the primary driver of convergence degradation, whereas FP4 in Fprop and Dgrad alone introduces only modest additional token requirements. To interpret this behavior, we evaluate both structured and stochastic interventions under a controlled experimental setting. We find that stochastic rounding and randomized Hadamard rotations fail to stabilize training once Wgrad is quantized, whereas deterministic Hadamard rotations consistently restore stable optimization. These results suggest that FP4 training instability is driven by structured micro-scaling errors along sensitive gradient paths, rather than by insufficient stochasticity. We run experiments with native MXFP4 support on AMD Instinct MI355X GPUs, enabling controlled investigation of these effects without reliance on software emulation. --- ### Moir: Let the Model Direct Its Own Story for Robust Cross-Domain Knowledge Editing Authors: Jea Kwon, Jiwon Kim, Dong-kyum Kim, Meeyoung Cha Published: 2026-05-10 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2607.20433v1 While language models remain frozen at their training state, the world evolves continuously. Knowledge editing has emerged as a key alternative to full retraining, but its deployment is bottlenecked by the erosion of core capabilities: mathematical and programmatic reasoning collapse while encyclopedic recall remains intact. We trace this asymmetric degradation to a distributional mismatch. Covariance-based editors preserve only the subspaces spanned by their reference corpus, but fail to capture the operative distribution shaped by post-training such as SFT and DPO. Static external corpora, including Wikipedia and even the original pretraining mixture, cannot recover this shifted manifold. We propose Moir, which estimates the preservation covariance $C$ directly from the model itself by sampling from its own decoding distribution. Seeding generation with a single random vocabulary token bypasses the instruction-following templates that otherwise dominate sampled outputs, exposing the broader subspaces the model has internalized. Moir requires no external data and serves as a drop-in component for any covariance-based editor, a practical advantage given that the pre- and post-training corpora of most modern LLMs are not publicly accessible. Across OLMo-2, Llama-3.1, and Qwen-3 (7-8B), under both MEMIT and AlphaEdit and in batch and sequential regimes, Moir consistently extends preservation in the most vulnerable domains, most strikingly on Qwen3-8B after 20,000 AlphaEdit batch edits, it retains 79.9% GSM8K accuracy compared to 10.9% with the Wikipedia baseline. These results suggest that aligning the preservation distribution with the model's operative distribution is a key factor in non-destructive editing, and that the model itself may be the most accessible source of that distribution for deployed systems. --- ### Position: Natural Language Should Not Fully Replace Formal Languages Authors: Eitan Wagner, Elisha Rosensweig, Omri Abend Published: 2026-05-10 Categories: cs.CL, cs.PL arXiv: https://arxiv.org/abs/2607.20432v1 Recent advances in large language models and their widespread adoption have prompted claims that natural language could entirely replace formal languages, such as programming languages for software design. In this position paper, we argue that this perspective overlooks fundamental linguistic properties of natural language, specifically that it is optimized for underspecification in open-ended contexts. We introduce a formal framework centered on *task specificity*, defining it as the information-theoretic reduction of uncertainty in an output space -- such as all possible images -- given a user's specific requirements. We prove a *specificity crossover theorem*, showing the existence of a threshold beyond which the cost to express formal requirements into natural language exceeds the cost of direct formal specification. By analyzing case studies across modalities, such as image generation, code synthesis, and audio production, we demonstrate that natural language excels at low specificity tasks, while formal languages are advantageous on tasks with stricter requirements. We conclude that natural and formal languages are complementary tools and advocate the development of hybrid systems that allow users to move across the specificity spectrum. --- ### Skill-Contracted Agents for Evidence-Aware Materials Literature Analysis Authors: Bixuan Li, Yu Liu, Shuo Shi, Xiaoya Huang, Peng Kang, Lei Zheng Published: 2026-05-10 Categories: cs.CL, cs.IR arXiv: https://arxiv.org/abs/2607.20431v1 Materials science literature analysis requires simultaneous attention to composition, processing, characterization, and property relationships, yet conventional retrieval-augmented generation pipelines struggle to reconcile heterogeneous tasks within a single retrieve-then-generate architecture. Here we present AlphaAgent, a skill-driven agent framework that decouples retrieval-based question answering from paper-level report generation through explicit skill contracts. A dedicated retrieval skill rewrites user requests into material-specific search intents, queries a curated index of more than 300,000 papers from the Journal Citation Reports Metallurgy and Metallurgical Engineering category, and reformulates queries when initial evidence is insufficient. A separate report-generation skill parses full-text PDFs to produce structured per-paper analytical reports and cross-paper summaries. In a blind evaluation on 40 materials-science questions, half of which required deep analytical reasoning, AlphaAgent substantially outperformed a baseline system matched for underlying model, document index, and retrieval scale, with the largest gains in mechanistic explanation and awareness of credibility boundaries. These results indicate that explicit task separation, refined retrieval intent, and evidence-aware generation improve large-language-model-based literature analysis for materials research. --- ### LLM-INSTRUCT at UZH Shared Task 2026: Constraint-Aware Retrieval and Selective Debate for Paragraph-Level Argument Mining Authors: Phuong Huu Vu Tran, Long Minh Vo, Son Nguyen Minh Le, Hoang Van Published: 2026-05-10 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2607.20430v1 We present LLM-INSTRUCT, the winning system for the UZH Shared Task at ArgMining 2026 on paragraph-level argument mining in UN and UNESCO resolutions. The task requires paragraph-type classification, prediction of a subset of 141 official tags, and directed relation prediction under a strict JSON schema setting using only open-weight models up to 8B parameters. We frame the task as constrained structured prediction. The system first narrows the candidate tag space with metadata-aware dense retrieval, then applies constrained decoding with per-dimension caps, escalates only uncertain cases to a three-agent debate branch, and finally validates the output schema. On the official leaderboard, LLM-INSTRUCT ranked 1st overall, with 1st in F1 and 5th in LLM-as-a-Judge. During development, our configuration search further improved Task 1b Micro-F1 from 35.83% to 40.08% while keeping the internal Task 2 score at 4.421. The main lesson is simple: reducing the decision space before generation improves both accuracy and submission robustness. Our code and supporting scripts are publicly available at: https://github.com/LLM-Instruct-at-UZH-Shared-Task-2026/Method --- ### More Is Not More: What Matters for Diversity in LLM Opinions? Authors: Qiyang Yao Published: 2026-05-10 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2607.20429v1 Large language models are increasingly used to simulate diverse human opinions in open-ended tasks such as synthetic surveys, focus group modeling, and public opinion prediction. However, LLM outputs exhibit systematic opinion homogenization. Practitioners have explored various interventions to increase diversity, but the landscape remains fragmented: different methods are evaluated in isolation with incomparable metrics, and in practice they are typically deployed and upgraded simultaneously, making it difficult to attribute gains to specific components. To advance a more scientific understanding of LLM output diversity, we design a factorial experiment that separates two primary intervention dimensions: input conditioning (operationalized through persona depth) and interaction architecture. We evaluate all conditions on 100 real-user open-ended questions across 7 models, measuring diversity with multiple complementary metrics. Our findings challenge several common assumptions. First, more persona detail does not monotonically increase diversity. The initial step of persona conditioning already captures the majority of the gain, while further elaboration with demographic detail does not consistently improve and can reduce diversity on some models. Second, rather than seeking a single best interaction architecture, we find that different architectures explore largely non-overlapping opinion regions. Combining multiple architectures yields broader coverage than optimizing any one. Third, commonly attempted low-cost alternatives such as raising sampling temperature and adding diversity instructions produce negligible effects compared to structured interventions. Overall, our work demonstrates that diversity is not a product of scaling along any single dimension, but is highly sensitive to the structural form and combination of interventions. --- ### Integro-differential equations in angular stabilization of drone motion by distributed feedback control Authors: Alexander Domoshnitsky, Oleg Kupervasser, Anatoly Polonsky Published: 2026-05-10 Categories: cs.AI, cs.LG, math.OC arXiv: https://arxiv.org/abs/2607.18251v1 In this paper, we propose angular stabilization of drone motion using distributed feedback control in the form of an integral operator. It should be stressed that the memory of this integral operator could be unbounded. It is intuitively clear that large length of the observation time open new possibilities to construct better control based on previous states of the control object. Unbounded memory in control requires the creation of a certain approach different from standard ones to the study of integro-differential equations. One of the goals of this article is to propose a certain universal approach that allows us to study the stability of integro-differential equations in the case of unbounded memory in the integral operator specifying the feedback control in stabilization. The approach we propose allows us to reduce the study of integro-differential equations to the analysis of systems of ordinary differential equations. In general, such systems can consist of an infinite number of equations. In relation to the so-called linear approximation in the problem of angle stabilization manages to limit itself to relatively simple exponential kernels in the integral control and arrive at a system with a finite number of equations. The examples explain that more complex kernels, for example, linear combinations of the exponential kernels, can enhance the stabilization capabilities. We obtain new unexpectable results on the exponential stability of integro-differential equations. Then we apply them to stabilization of drone flight. --- ### Shapley Context Pruning: A Cooperative Game Perspective for Context Reranking and Pruning Authors: Yanqiao Chen, Dongsheng Hou, Yuhan Rui, Zhen Cao, Yepang Liu Published: 2026-05-10 Categories: cs.AI arXiv: https://arxiv.org/abs/2607.16209v1 Context reranking and pruning have become essential for improving the efficiency of modern Retrieval-Augmented Generation (RAG) systems, yet an interpretable and unified framework remains underexplored. Previous work has primarily emphasized lexical retrieval, cross-encoder architectures, model distillation, and Low-Rank Adaptation (LoRA), mostly relying on heuristic loss functions and empirical attribution. This paper presents Shapley Context Pruning (SCP), a novel framework for context reranking that establishes a cooperative-game-theory perspective for importance attribution by modeling the context as a cooperative game. Balancing the trade-off between fine-grained and coarse-grained representations, we employ a Deep Sets architecture to approximate a permutation-invariant value function at the sentence level, utilizing pre-trained language models as sentence embedders and optimizing via a pairwise margin ranking loss. To ensure practical scalability without sacrificing mathematical rigor, we leverage Monte-Carlo sampling for efficient training and inference, providing formal theoretical error bounds and sample complexity guarantees for preserving Top-K subset rankings. Furthermore, we conduct comprehensive experiments-spanning supporting-sentence recall, Needle-in-the-Haystack (NIAH) evaluations, long-context QA, and multi-hop reasoning-alongside rigorous ablation studies on embedding quality and attribution strategies. The model achieves competitive downstream QA performance against robust baselines. ---