# Research Papers — 2026-05 · bundle page 107 of 326 > Full-text markdown of papers #5301–5350 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/107 [← Previous bundle](https://aipod.dev/papers/archive/2026-05/page/106.md) [Next bundle →](https://aipod.dev/papers/archive/2026-05/page/108.md) ### MotiMotion: Motion-Controlled Video Generation with Visual Reasoning Authors: Lee Hsin-Ying, Hanwen Jiang, Yiqun Mei, Jing Shi, Ming-Hsuan Yang, Zhixin Shu Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22818v1 Current motion-controlled image-to-video generation models rigidly follow user-provided trajectories that are often sparse, imprecise, and causally incomplete. Such reliance often yields unnatural or implausible outcomes, especially by missing secondary causal consequences. To address this, we introduce MotiMotion, a novel framework that reformulates motion control as a reasoning-then-generation problem. To encourage causally grounded and commonsense-consistent interactions, we leverage a training-free vision-language reasoner to refine image-space coordinates of primary trajectories and to hallucinate plausible secondary motions. To further improve motion naturalness, we propose a confidence-aware control scheme that modulates guidance strength, enabling the model to closely follow high-confidence plans while correcting artifacts under low-confidence inputs with its internal generative priors. To support systematic evaluation, we curate a new image-to-video benchmark, MotiBench, consisting of interaction-centric scenes where new events are triggered by motion. Both VLM-based evaluation and a human study on MotiBench demonstrate that MotiMotion produces videos with more plausible object behaviors and interaction, and is preferred over existing approaches. --- ### Vector Policy Optimization: Training for Diversity Improves Test-Time Search Authors: Ryan Bahlous-Boldi, Isha Puri, Idan Shenfeld, Akarsh Kumar, Mehul Damani, Sebastian Risi, Omar Khattab, Zhang-Wei Hong, Pulkit Agrawal Published: 2026-05-21 Categories: cs.LG, cs.AI, cs.CL, cs.NE arXiv: https://arxiv.org/abs/2605.22817v1 Language models must now generalize out of the box to novel environments and work inside inference-scaling search procedures, such as AlphaEvolve, that select rollouts with a variety of task-specific reward functions. Unfortunately, the standard paradigm of LLM post-training optimizes a pre-specified scalar reward, often leading current LLMs to produce low-entropy response distributions and thus to struggle at displaying the diversity that inference-time search will require. We propose Vector Policy Optimization (VPO), an RL algorithm that explicitly trains policies to anticipate diverse downstream reward functions and to produce diverse solutions. VPO exploits that rewards are often vector-valued in practice, like per-test-case correctness in code generation or, say, multiple different user personas or reward models. VPO is essentially a drop-in replacement for the GRPO advantage estimator, but it trains the LLM to output a set of solutions where individual solutions specialize to different trade-offs in the vector reward space. Across four tasks, VPO matches or beats the strongest scalar RL baselines on test-time search (e.g. pass@k and best@k), with the gap widening as the search budget grows. For evolutionary search, VPO models unlock problems that GRPO models cannot solve at all. As test-time search becomes more standardized, optimizing for diversity may need to become the default post-training objective. --- ### AwareVLN: Reasoning with Self-awareness for Vision-Language Navigation Authors: Wenxuan Guo, Xiuwei Xu, Yichen Liu, Xiangyu Li, Hang Yin, Huangxing Chen, Wenzhao Zheng, Jianjiang Feng, Jie Zhou, Jiwen Lu Published: 2026-05-21 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2605.22816v1 Vision-and-Language Navigation (VLN) requires an agent to ground language instructions to its own movement within a visual environment. While state-of-the-art methods leverage the reasoning capabilities of Vision-Language Models (VLMs) for end-to-end action prediction, they often lack an explicit and explainable understanding of the relationships between the agent, the instruction, and the scene. Conversely, explicitly building a scene map for heuristic planning is intuitively appealing but relies on additional 3D sensors and hinders large-scale vision-language pre-training. To bridge this gap, we propose AwareVLN, a novel framework that equips the navigation model with a self-aware reasoning mechanism, enabling it to understand the agent's state and task progress in a fully end-to-end and data-driven manner. Our approach features two key innovations: (1) a structural reasoning module that fosters spatial and task-oriented self-awareness, and (2) an automatic data engine with progress division for effective training. Extensive experiments on various datasets in Habitat simulator show our AwareVLN significantly outperforms previous state-of-the-art vision-language navigation methods. Project page: https://gwxuan.github.io/AwareVLN/. --- ### Remember to be Curious: Episodic Context and Persistent Worlds for 3D Exploration Authors: Lily Goli, Justin Kerr, Daniele Reda, Alec Jacobson, Andrea Tagliasacchi, Angjoo Kanazawa Published: 2026-05-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.22814v1 Exploration is a prerequisite for learning useful behaviors in sparse-reward, long-horizon tasks, particularly within 3D environments. Curiosity-driven reinforcement learning addresses this via intrinsic rewards derived from the mismatch between the agent's predictive model of the world and reality. However, translating this intrinsic motivation to complex, photorealistic environments remains difficult, as agents can become trapped in local loops and receive fresh rewards for revisiting forgotten states. In this work, we demonstrate that this failure stems from a lack of spatial persistence and episodic context. We show that effective curiosity requires a model of the world that is persistent and continuously updated, paired with an agent that maintains an episodic trajectory history to navigate toward novel regions. We achieve this using an online 3D reconstruction as a persistent model of the world, while the agent policy is parameterized as a sequence model over RGB observations to maintain episodic context. This design enables effective exploration during training while allowing the agent to navigate using solely RGB frames at deployment. Trained purely via curiosity on HM3D, our agent outperforms RL-based active mapping baselines and generalizes zero-shot to Gibson and AI-generated worlds. Our end-to-end policy enables efficient adaptation to downstream tasks, such as apple picking and image-goal navigation, outperforming from-scratch baselines. Please see video results at https://recuriosity.github.io/. --- ### GesVLA: Gesture-Aware Vision-Language-Action Model Embedded Representations Authors: Wenxuan Guo, Ziyuan Li, Meng Zhang, Yichen Liu, Yimeng Dong, Chuxi Xu, Yunfei Wei, Ze Chen, Erjin Zhou, Jianjiang Feng Published: 2026-05-21 Categories: cs.RO, cs.CV arXiv: https://arxiv.org/abs/2605.22812v1 Vision-Language-Action (VLA) models have shown strong potential for general-purpose robot manipulation by unifying perception and action. However, existing VLA systems primarily rely on textual instructions and struggle to resolve spatial ambiguity in complex scenes with multiple similar objects. To address this limitation, we introduce gesture as a parallel instruction modality and propose a Gesture-aware Vision-Language-Action model (GesVLA). Our approach encodes gesture features directly into the latent space, enabling them to participate in both high-level reasoning and low-level action generation, and adopts a dual-VLM architecture to achieve tight coupling between gesture representations and action policies. At the data level, we construct a scalable gesture data generation pipeline by rendering hand models onto real-world scene images. This reduces the sim-to-real visual gap while producing rich data with diverse motion patterns and corresponding pointing annotations. In addition, we employ a two-stage training strategy to equip the model with both gesture perception and action prediction capabilities. We evaluate our approach on multiple real-world robotic tasks, including a controlled block manipulation task for validation and more practical scenarios such as product and produce selection. Experimental results show that incorporating gesture consistently improves target grounding accuracy and human-robot interaction efficiency, especially in complex and cluttered environments. Project page: https://gwxuan.github.io/GesVLA/. --- ### Sensor2Sensor: Cross-Embodiment Sensor Conversion for Autonomous Driving Authors: Jiahao Wang, Bo Sun, Yijing Bai, Vincent Casser, Songyou Peng, Zehao Zhu, Meng-Li Shih, Xander Masotto, Shih-Yang Su, Kanaad V Parvate, Tiancheng Ge, Linn Bieske et al. Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22809v2 Robust training and validation of Autonomous Driving Systems (ADS) require massive, diverse datasets. Proprietary data collected by Autonomous Vehicle (AV) fleets, while high-fidelity, are limited in scale, diversity of sensor configurations, as well as geographic and long-tail-behavioral coverage. In contrast, in-the-wild data from sources like dashcams offers immense scale and diversity, capturing critical long-tail scenarios and novel environments. However, this unstructured, in-the-wild video data is incompatible with ADS expecting structured, multi-modal sensor inputs for validation and training. To bridge this data gap, we propose Sensor2Sensor, a novel generative modeling paradigm that translates in-the-wild monocular dashcam videos into a high-fidelity, multi-modal sensor suite (AV logs) comprising multi-view camera images and LiDAR point clouds. A core challenge is the lack of paired training data. We address this by converting real AV logs into dashcam-style videos via 4D Gaussian Splatting (4DGS) reconstruction and novel-view rendering. Sensor2Sensor then utilizes a diffusion architecture to perform the generative conversion. We perform comprehensive quantitative evaluations on the fidelity and realism of the generated sensor data. We demonstrate Sensor2Sensor's practical utility by converting challenging in-the-wild internet and dashcam footage into realistic, multi-modal data formats, further unlocking vast external data sources for AV development. --- ### The Matching Principle: When Does a Training Penalty Cover Deployment Shift? Authors: Vishal Rajput Published: 2026-05-21 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2605.22800v3 Ordinary training optimises the task loss and then stops. It never pays for internal representation energy: Jacobians can stay large in directions that never helped the label, so even small label-preserving noise throws the model off---a design gap that classical noise-injection theory fixes at second order, but only when applied as default regularisation, which current practice does not do. We make that precise with a Matching Principle: name deployment directions (Sigma_task) and the training penalty Sigma', and ask whether the second covers the first. The no-thinking default is even-spread / isotropic penalty (Sigma' proportional to I)---classical Gaussian / Tikhonov at second order: no axis estimate, no architecture change, and---in a simple linear ridge model---strictly less deployment drift than task-only training, with no coverage miss by construction. When axes are known, matching is sharper; when they are missed, a residual floor remains. Across seven domains a named second-moment penalty beats unregularised training; a controlled illustration recovers match > even-spread > wrong-axis when axes are forced. The ridge theorems are proved; deep nets remain experiments under a specified perturbation. Design rule: fix internal energy by default (even-spread); match when axes are known; treat losses that control representation sensitivity as first-class design. --- ### The Matching Principle: A Geometric Theory of Loss Functions for Nuisance-Robust Representation Learning Authors: Vishal Rajput Published: 2026-05-21 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2605.22800v2 Robustness, domain adaptation, photometric/occlusion invariance, sensor drift, and alignment style are treated as separate literatures with separate method families. Under label-preserving deployment shift they share one geometric object: the covariance Sigma_task = Cov_{Q_n}(n) of ways inputs can change without changing the label. CORAL, adversarial training, augmentation, metric learning, Jacobian penalties, and alignment constraints are not independent tricks--they are estimators of Sigma_task. Fix that object and the Jacobian penalty is pinned by a matrix Sigma' whose range must cover range(Sigma_task)--the matching principle. We prove optimality in a linear-Gaussian model (Thm. A), necessity of range coverage for any quadratic penalty that zeros deployment drift (Thm. G), and the same dichotomy at global minima (Thm. A*_global). Wrong-direction/signal-aligned controls (Lemma C; Cor. E/E*) and seven estimators (Lemmas D1--D7), plus label-free TDI, yield a falsifiable recipe when Sigma_task must be learned. Thirteen blocks (ML through Qwen2.5-7B) test matched vs isotropic vs wrong-direction penalties on geometry and deployment drift. Twelve match theory where identifiability holds; Office-31 is a named eigengap failure. Partial passes: geometry can improve without every headline task metric moving. A pilot 7B DPO run (one epoch, 240 pairs): matched style-PMH preserves Style TDI where standard DPO degrades it. We do not claim standard training reaches global minima (assumption (O) is open), that estimated Sigma_task is always identifiable, or dominance on every leaderboard. We claim a falsifiable design recipe: estimate Sigma_task, match Sigma', run the controls, report task and geometry separately. --- ### Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models Authors: Krishnakumar Balasubramanian Published: 2026-05-21 Categories: stat.ML, cs.AI, cs.LG, math.ST arXiv: https://arxiv.org/abs/2605.22795v3 We analyze finite-particle drifting models for one-step generative modeling. For a conservative velocity given by the difference of the kernel-smoothed data and model scores, a joint-entropy identity yields continuous-time bounds for the smoothed Fisher discrepancy and the squared particle velocity. The finite-particle correction involves reciprocal kernel density estimates. We give local-occupancy conditions and exact expectation bounds, and show that these expectations diverge for full-support initial densities in a shrinking-bandwidth regime. Keeping the bandwidth dependence of the quadrature constants explicit yields a conditional root residual-velocity rate of $N^{-1/(d+4)}$ under uniform regularity, with a corresponding rate under weaker quadrature growth conditions. We also analyze the original displacement field with the exact Laplace kernel. A companion kernel gives a weighted coercivity estimate that accommodates the unbounded empirical scale and the kernel's nondifferentiability. Localization controls the particle velocity, while relative scale alignment absorbs the mismatch into dissipation. Both analyses quantify the size of an additional drift correction under explicit regularity and stability assumptions. --- ### Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models Authors: Krishnakumar Balasubramanian Published: 2026-05-21 Categories: stat.ML, cs.AI, cs.LG, math.ST arXiv: https://arxiv.org/abs/2605.22795v2 We propose and analyze a conservative drifting method for one-step generative modeling. The method replaces the original displacement-based drifting velocity by a kernel density estimator (KDE)-gradient velocity, namely the difference of the kernel-smoothed data score and the kernel-smoothed model score. This velocity is a gradient field, addressing the non-conservatism issue identified for general displacement-based drifting fields. We prove continuous-time finite-particle convergence bounds for the conservative method on $\R^d$: a joint-entropy identity yields bounds for the empirical Stein drift, the smoothed Fisher discrepancy of the KDE, and the squared center velocity. The main finite-particle correction is a reciprocal-KDE self-interaction term, and we give deterministic and high-probability local-occupancy conditions under which this term is controlled. We keep the quadrature constants explicit and track their possible bandwidth dependence: the root residual-velocity rate $N^{-1/(d+4)}$ holds under an additional $h$-uniform quadrature regularity condition, while a more general growth condition yields the optimized root rate $N^{-(2-β)/(2(d+4-β))}$, where $0\le β<2$. We also analyze the non-conservative drifting method with Laplace kernel, corresponding to the original displacement-based velocity proposed in Deng et al., 2026 (arxiv:2602.04770). For this method, a sharp companion kernel decomposes the velocity into a positive scalar preconditioning of a sharp-score mismatch plus a Laplace scale-mismatch residual, producing an analogous finite-particle rate with an unavoidable residual term. Finally, we explain how the continuous-time residual-velocity bounds translate into one-step generation guarantees through the explicit drift size $η$. --- ### MOSS: Self-Evolution through Source-Level Rewriting in Autonomous Agent Systems Authors: Qianshu Cai, Yonggang Zhang, Xianzhang Jia, Huajiang Zheng, Wei Xue, Jun Song, Xinmei Tian, Yike Guo Published: 2026-05-21 Categories: cs.AI, cs.LG arXiv: https://arxiv.org/abs/2605.22794v2 Autonomous agentic systems are largely static after deployment: they do not learn from user interactions, and recurring failures persist until the next human-driven update ships a fix. Self-evolving agents have emerged in response, but all confine evolution to text-mutable artifacts -- skill files, prompt configurations, memory schemas, workflow graphs -- and leave the agent harness untouched. Since routing, hook ordering, state invariants, and dispatch live in code rather than in any text artifact, an entire class of structural failure is physically unreachable from the text layer. We argue that source-level adaptation is a fundamentally more general medium: it is Turing-complete, a strict superset of every text-mutable scope, takes effect deterministically rather than through base-model compliance, and does not erode under long-context drift. We present MOSS, a system that performs self-rewriting at the source level on production agentic substrates. Each evolution is anchored to an automatically curated batch of production-failure evidence and proceeds through a deterministic multi-stage pipeline; code modification is delegated to a pluggable external coding-agent CLI while MOSS retains stage ordering and verdicts. Candidates are verified by replaying the batch against the candidate image in ephemeral trial workers, then promoted via user-consent-gated, in-place container swap with health-probe-gated rollback. On OpenClaw, MOSS lifts a four-task mean grader score from 0.25 to 0.61 in a single cycle without human intervention. --- ### Gated DeltaNet-2: Decoupling Erase and Write in Linear Attention Authors: Ali Hatamizadeh, Yejin Choi, Jan Kautz Published: 2026-05-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.22791v1 Linear attention replaces the unbounded cache of softmax attention with a fixed-size recurrent state, reducing sequence mixing to linear time and decoding to constant memory. The hard part is not just what to forget, but how to edit this compressed memory without scrambling existing associations. Delta-rule models subtract the current read before writing a new value, and Kimi Delta Attention (KDA) sharpens forgetting with channel-wise decay. But the active edit still uses a single scalar gate to control two different things: how much old content to erase on the key side and how much new content to commit on the value side. We introduce Gated DeltaNet-2, which generalizes both Gated DeltaNet and KDA by inheriting adaptive forgetting and channel-wise decay while addressing their shared limitation, the scalar tie between erasing and writing. Gated Delta Rule-2 separates these roles with a channel-wise erase gate b_t and a channel-wise write gate w_t, reducing to KDA when both gates collapse to the same scalar and to Gated DeltaNet when the decay also collapses. We derive a fast-weight update view, a chunkwise WY algorithm with channel-wise decay absorbed into asymmetric erase factors, and a gate-aware backward pass that preserves efficient parallel training. At 1.3B parameters trained on 100B FineWeb-Edu tokens, Gated DeltaNet-2 achieves the strongest overall results among Mamba-2, Gated DeltaNet, KDA, and Mamba-3 variants across language modeling, commonsense reasoning, and retrieval. Its advantage is most pronounced on long-context RULER needle-in-a-haystack benchmarks, where it improves the evaluated multi-key retrieval setting and remains strong in both recurrent and hybrid settings. Code is available at https://github.com/NVlabs/GatedDeltaNet-2. --- ### LCGuard: Latent Communication Guard for Safe KV Sharing in Multi-Agent Systems Authors: Sadia Asif, Mohammad Mohammadi Amiri, Momin Abbas, Prasanna Sattigeri, Karthikeyan Natesan Ramamurthy Published: 2026-05-21 Categories: cs.AI, cs.ET, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2605.22786v1 Large language model (LLM)-based multi-agent systems increasingly rely on intermediate communication to coordinate complex tasks. While most existing systems communicate through natural language, recent work shows that latent communication, particularly through transformer key-value (KV) caches, can improve efficiency and preserve richer task-relevant information. However, KV caches also encode contextual inputs, intermediate reasoning states, and agent-specific information, creating an opaque channel through which sensitive content may propagate across agents without explicit textual disclosure. To address this, we introduce \textbf{LCGuard} (Latent Communication Guard), a framework for safe KV-based latent communication in multi-agent LLM systems. LCGuard treats shared KV caches as latent working memory and learns representation-level transformations before cache artifacts are transmitted across agents. We formalize representation-level sensitive information leakage operationally through reconstruction: a shared cache artifact is unsafe if an adversarial decoder can recover agent-specific sensitive inputs from it. This leads to an adversarial training formulation in which the adversary learns to reconstruct sensitive inputs, while LCGuard learns transformations that preserve task-relevant semantics and reduce reconstructable information. Empirical evaluations across multiple model families and multi-agent benchmarks show that LCGuard consistently reduces reconstruction-based leakage and attack success rates while maintaining competitive task performance compared to standard KV-sharing baselines. --- ### Evaluating Commercial AI Chatbots as News Intermediaries Authors: Mirac Suzgun, Emily Shen, Federico Bianchi, Alexander Spangher, Thomas Icard, Daniel E. Ho, Dan Jurafsky, James Zou Published: 2026-05-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2605.22785v1 AI chatbots are rapidly shaping how people encounter the news, yet no prior study has systematically measured how accurately these systems, with their proprietary search integrations and retrieval-synthesis pipelines, handle emerging facts across languages and regions. We present a 14-day (February 9-22, 2026) evaluation of six AI chatbots (Gemini 3 Flash and Pro, Grok 4, Claude 4.5 Sonnet, GPT-5 and GPT-4o mini) on 2,100 factual questions derived from same-day BBC News reporting across six regional services (US & Canada, Arabic, Afrique, Hindi, Russian, Turkish). The best systems achieve over 90% multiple-choice accuracy on questions about events reported hours earlier. The same systems, however, lose 11-13% under free-response evaluation, and 16-17% across the cohort. We further characterize three failure patterns. First, every model achieves its lowest accuracy on Hindi (79% vs. 89-91% elsewhere) and citations indicate an Anglophone retrieval bias (e.g., models answering Hindi queries cite English Wikipedia more than any Hindi outlet). Second, retrieval, not reasoning, failures drive over 70% of all errors. When models retrieve a correct source, they often extract the correct answer; the problem is to land on the right source in the first place. Third, models achieving 88-96% accuracy on well-formed questions drop to 19-70% when questions contain subtle false premises, with the most vulnerable model accepting fabricated facts 64% of the time. We also identify a detection-accuracy paradox: the best false-premise detector ranks second in adversarial accuracy (abstention rate), while a weaker detector ranks first, showing that premise detection and answer recovery are partially independent capabilities. Overall, these suggest that high accuracy can mask systematic regional inequity, near-total dependence on retrieval infrastructure, and vulnerability to imperfect queries real users pose. --- ### DeltaBox: Scaling Stateful AI Agents with Millisecond-Level Sandbox Checkpoint/Rollback Authors: Yunpeng Dong, Jingkai He, Shiqi Liu, Yuze Hou, Dong Du, Zhonghu Xu, Si Yu, Baochuan Yang, Yubin Xia, Haibo Chen Published: 2026-05-21 Categories: cs.OS, cs.AI arXiv: https://arxiv.org/abs/2605.22781v2 LLM-powered AI agents require high-frequency state exploration (e.g., test-time tree search and reinforcement learning), relying on rapid checkpoint and rollback (C/R) of the complete sandbox state, including files and process state (e.g., memory, contexts, etc.). Existing mechanisms duplicate the entire state, causing hundreds of milliseconds to seconds of latency per C/R, which severely bottlenecks deep search and large-scale fan-outs. This paper observes that subsequent checkpoints in AI agents are highly similar. Therefore, instead of full duplication, a sandbox should only duplicate the changes between consecutive checkpoints (Key Insight). However, it is non-trivial to realize the idea, mainly due to the missing OS supports. This paper proposes a new OS-level abstraction, DeltaState, to enable the change-based transactional C/R for AI agents with two co-designed OS mechanisms. First, DeltaFS enables change-based filesystem C/R by organizing the file states into layers and dynamically freezing the writable layer and inserting a new one during checkpoint, reducing file updates to copy-on-write, and making rollback a simple layer switch. Second, DeltaCR enables change-based process state C/R using incremental dumps, and accelerates rollback by bypassing traditional pipelines to directly fork() from a frozen template process. We then present DeltaBox, a novel agent sandbox achieving millisecond level C/R through the two new mechanisms. Evaluations on SWE-bench and RL micro-benchmarks show DeltaBox completes checkpoint and rollback in millisecond-level latency (14ms and 5ms, respectively), empowering agents to explore substantially more nodes under fixed time budgets. --- ### FAME: Failure-Aware Mixture-of-Experts for Message-Level Log Anomaly Detection Authors: Huanchi Wang, Zihang Huang, Yifang Tian, Kristina Dzeparoska, Hans-Arno Jacobsen, Alberto Leon-Garcia Published: 2026-05-21 Categories: cs.SE, cs.LG arXiv: https://arxiv.org/abs/2605.22779v2 Production systems generate millions of log lines daily, yet most anomaly detectors operate at the session or window-level, flagging groups of lines rather than identifying the specific message responsible. This coarse granularity forces operators to inspect many routine lines per alert. Message-level detection offers finer granularity, but remains challenging. A single event template may correspond to both normal and anomalous messages, failures arise from heterogeneous subsystems, and line-level labeling at scale is impractical. Although large language models (LLMs) can reason over log semantics, applying them to every line is too costly for continuous monitoring. We present FAME (Failure-Aware Mixture-of-Experts), a label-efficient message-level mixture-of-experts framework that uses an LLM only once offline. We annotate at most K labeled lines per template to derive binary normal/anomaly indicators and representative examples. The LLM proposes a partition of templates into failure domains, and a certification step validates the proposal before training. FAME trains a lightweight router and domain experts that run on-premise and output anomaly predictions and failure-domain labels. On BGL, FAME achieves F1 = 98.16 at K = 100 reducing annotation effort by 76x and detects 97.7% of anomalies from unseen EventIDs. On Thunderbird, FAME reaches F1 = 99.95 with perfect recall. --- ### DecQ: Detail-Condensing Queries for Enhanced Reconstruction and Generation in Representation Autoencoders Authors: Tianhang Wang, Yitong Chen, Wei Song, Zuxuan Wu, Min Li, Jiaqi Wang Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22777v1 Representation Autoencoders (RAEs) leverage frozen vision foundation models (VFMs) as tokenizer encoders, providing robust high-level representations that facilitate fast convergence and high-quality generation in latent diffusion models. However, freezing the VFM inherently constrains its spatial reconstruction capacity, limiting fine-grained generation and image editing; in contrast, incorporating reconstruction-oriented signals via fine-tuning disrupts the pretrained semantic space and degrades generative fidelity. To address this trade-off, we propose DecQ, a simple yet effective framework for RAEs. Specifically, DecQ introduces lightweight detail-condensing queries that extract fine-grained information from intermediate VFM features through condenser modules. These queries are incorporated into the decoder to support reconstruction and are jointly generated with patch tokens during generative modeling. By aggregating information from both shallow and deep layers, DecQ effectively mitigates the reconstruction--generation trade-off, improving both reconstruction quality and generative performance. Our experiments demonstrate that: (1) with only 8 additional queries and 3.9% extra computation, DecQ improves reconstruction over the frozen DINOv2-based RAE, increasing PSNR from 19.13 dB to 22.76 dB; and (2) for generative modeling, DecQ achieves 3.3$\times$ faster convergence than RAE, attaining an FID of 1.41 without guidance and 1.05 with guidance. --- ### SDPM: Survival Diffusion Probabilistic Model for Continuous-Time Survival Analysis Authors: Stanislav R. Kirpichenko, Andrei V. Konstantinov, Lev V. Utkin Published: 2026-05-21 Categories: cs.LG, cs.AI, stat.CO, stat.ML arXiv: https://arxiv.org/abs/2605.22776v1 Survival analysis aims to estimate a time-to-event distribution from data with censored observations. Many existing methods either impose structural assumptions on the hazard function or discretize the time axis, which may limit flexibility and introduce approximation errors. We propose the Survival Diffusion Probabilistic Model (SDPM), a generative approach to continuous-time survival analysis. SDPM models the conditional distribution of the survival outcome, represented by the pair of observed time and censoring indicator, $\mathbb{P}(T,δ\mid \mathbf{x})$, using a denoising diffusion model. Under the assumption of conditionally independent censoring, conditional samples generated by the model can be transformed into survival function estimates using the Kaplan-Meier estimator. This formulation avoids parametric assumptions on the event-time distribution and does not require a discretization of the output time space. The model operates in a transformed target space, using standardized log-times and a continuous Gaussian-mixture representation of the censoring indicator. We evaluate SDPM on ten real survival datasets and compare it with five strong baselines, including tree-based, boosting-based, and neural survival models. Results show that SDPM achieves competitive predictive performance across C-index, integrated time-dependent AUC, and integrated Brier score. A study on synthetic Cox-Weibull data demonstrates that SDPM can recover the shape of an underlying continuous survival distribution more accurately than a strong nonparametric baseline when sufficiently many samples are generated. An ablation study confirms the importance of the proposed target-space transformations, which improve event-rate calibration, reduce invalid generated times, and provide consistent gains in predictive discrimination. Codes implementing the proposed model are publicly available. --- ### MambaGaze: Bidirectional Mamba with Explicit Missing Data Modeling for Cognitive Load Assessment from Eye-Gaze Tracking Data Authors: Amir Mousavi, Mohammad Sadegh Sirjani, Erfan Nourbakhsh, Mimi Xie, Rocky Slavin, Leslie Neely, John Davis, John Quarles Published: 2026-05-21 Categories: cs.LG, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2605.22775v3 Real-time cognitive load assessment from eye-tracking signals could enable adaptive human-centered AI in safety-critical applications such as driver vigilance monitoring or automated flight deck assistance, yet two challenges persist: handling frequent data missingness from blinks and tracking failures, and efficiently modeling long-range temporal dependencies. We propose MambaGaze (Bi-Mamba), a framework that addresses these challenges through (1) XMD encoding, which augments raw features with observation masks and time-deltas to explicitly model data uncertainty, and (2) bidirectional Mamba-2, which captures temporal dependencies with linear computational complexity. Experiments on CLARE and CL-Drive datasets under leave-one-subject-out evaluation show that MambaGaze achieves 77.1% accuracy and 59.2% macro-F1 on CLARE, and 69.4% accuracy and 51.5% macro-F1 on CL-Drive, attaining the highest average LOSO macro-F1 (55.3%) across all ten compared models. Input-stream ablation indicates that log-scaled time-deltas are the strongest single channel in our setting, and combining all three XMD streams provides consistent gains of 5-20 pp macro-F1. Edge deployment benchmarks on three NVIDIA Jetson Orin platforms show real-time inference at 27-36 FPS with power consumption below 6.6 W, supporting feasibility for embedded cognitive load monitoring. --- ### CogAdapt: Adapting Clinical ECG Foundation Models for Wearable Cognitive Load Assessment Authors: Amir Mousavi, Erfan Nourbakhsh, Mohammad Sadegh Sirjani, Mimi Xie, Rocky Slavin, Leslie Neely, John Davis, John Quarles Published: 2026-05-21 Categories: cs.LG, cs.AI, cs.HC arXiv: https://arxiv.org/abs/2605.22774v5 Assessing cognitive load continuously and at low latency would help adaptive human-computer interaction, but it remains hard because labeled data are scarce and models generalize poorly across subjects. Recent ECG foundation models, pre-trained on millions of clinical diagnostic ECG recordings, yet they do not apply directly to wearable devices when the sensor configuration and the task both differ. We present CogAdapt, a framework that adapts a clinical ECG foundation model to wearable cognitive load assessment. CogAdapt has two parts. LeadBridge is a learnable adapter that maps 3-lead wearable signals to a 12-lead-compatible representation. ProFine is a progressive fine-tuning strategy that unfreezes encoder layers in stages while limiting representational drift in the pre-trained model. On two public datasets (CLARE and CL-Drive) under leave-one-subject-out cross-validation, CogAdapt reaches macro-F1 of 0.626 and 0.768, improving over from-scratch baselines by 11.2 and 16.1 percentage points. The results show that a clinical ECG pretraining can support subject-independent cognitive load assessment from wearable sensors. --- ### Deep Reinforcement Learning for Flexible Job Shop Scheduling with Random Job Arrivals Authors: Yu Tang, Muhammad Zakwan, Efe Balta, John Lygeros, Alisa Rupenyan Published: 2026-05-21 Categories: cs.AI, math.OC arXiv: https://arxiv.org/abs/2605.22773v1 The Flexible Job Shop Scheduling Problem (FJSP) is the optimal allocation of a set of jobs to machines. Two primary challenges persist in FJSP: the unpredictable arrival of future jobs and the combinatorial complexity of the problem, rendering it intractable for conventional mixed-integer linear programming solvers. This paper proposes an event-based \gls{DRL} approach to solve FJSP with random job arrivals. Specifically, we employ the Proximal Policy Optimization algorithm and use lightweight Multi-Layer Perceptrons to train the \gls{DRL} agent for minimizing the total completion time of all jobs. We design the state representation to be directly accessible from the environment, and limit the learning agent to selecting from among a set of well-established dispatching rules. Simulations show that our \gls{DRL} approach outperforms any of the individual dispatching rules on datasets with varying heterogeneity and job arrival rates. We benchmark our \gls{DRL} against an arrival-triggered mixed-integer linear programming solution and show that our method achieves good performance especially when the datasets are heterogeneous. --- ### Reducing Political Manipulation with Consistency Training Authors: Long Phan, Devin Kim, Alexander Pan, Alice Blair, Adam Khoja, Dan Hendrycks Published: 2026-05-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2605.22771v2 Large language models (LLMs) exhibit systematic political bias across a variety of sensitive contexts. We find that LLMs handle counterpart topics from opposing political sides asymmetrically. We refer to this phenomenon as covert political bias and identify 7 categories of techniques through which it operates. We propose two metrics for covert bias: Sentiment Consistency measures symmetry in rhetoric and framing across paired political prompts; Helpfulness Consistency measures symmetric depth and engagement. To reduce both types of covert bias, we introduce Political Consistency Training (PCT), an RL training method with two complementary paradigms: Sentiment Consistency Training and Helpfulness Consistency Training. We show that PCT preserves overall helpfulness, substantially reduces covert political bias, and generalizes to held-out benchmarks. We release our work at https://political-manipulation.ai --- ### Understanding Data Temporality Impact on Large Language Models Pre-training Authors: Hippolyte Pilchen, Romain Fabre, Franck Signe Talla, Patrick Perez, Edouard Grave Published: 2026-05-21 Categories: cs.CL, cs.AI arXiv: https://arxiv.org/abs/2605.22769v2 Large language models (LLMs) are typically trained on shuffled corpora, yielding models whose knowledge is frozen at train time and whose temporal grounding remains poorly understood. In this work, we study the impact of pre-training dynamics on the acquisition of time-sensitive factual knowledge, focusing specifically on data ordering. Our main contributions are twofold. First, we introduce a comprehensive benchmark of over 7,000 temporally grounded questions and an evaluation protocol that enables analysis of whether models correctly associate facts with their corresponding time periods. Second, we pretrain 6B-parameter models on temporally ordered Common Crawl snapshots and compare them against standard shuffled pre-training. Our results show that sequentially trained models match shuffled baselines on general language understanding and common knowledge while consistently exhibiting more up-to-date and temporally precise knowledge. Temporally ordered pre-training yields improved factual freshness, while shuffled pre-training peaks on older data, possibly due to increased factual repetition. These findings, along with the release of our code at https://github.com/kyutai-labs/kairos , checkpoints, and datasets at https://huggingface.co/collections/kyutai/kairos provide a foundation for future research on continual learning for LLMs. --- ### Synthetic Data Alone is Enough? Rethinking Data Scarcity in Pediatric Rare Disease Recognition Authors: Ganlin Feng, Yuxi Long, Erin Lou, Lianghong Chen, Zihao Jing, Pingzhao Hu, Wei Xu Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22767v1 Children with rare genetic diseases often exhibit distinctive facial phenotypes, yet developing computer vision systems for early diagnosis remains challenging due to extreme data scarcity, privacy constraints, and limited data sharing in pediatric settings. These challenges not only hinder automated diagnosis but also restrict the availability of visual resources for clinical genetic counseling. While prior work has shown that synthetic data can augment real datasets and preserve phenotype-level semantics, it remains unclear whether synthetic data alone is sufficient for learning in ultra-low-resource pediatric settings. In this work, we study the synthetic-only regime for pediatric rare disease recognition. Under a controlled experimental setup, models are trained exclusively on phenotype-aware synthetic facial images at increasing scales. We find that synthetic-only training achieves performance comparable to real-data-only baselines at sufficient scale across multiple backbones, suggesting that high-fidelity synthetic data can approximate clinically meaningful distributions. These findings together further enable the use of synthetic pediatric facial images as privacy-preserving resources for genetic education and counseling, supporting clinician training and patient communication. Our results highlight the potential of computer vision to improve data efficiency and expand accessible visual tools in children's healthcare. --- ### Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation Authors: Samson Gourevitch, Yazid Janati, Dario Shariatian, Umut Simsekli, Eric Moulines, Eric P. Xing, Alain Durmus Published: 2026-05-21 Categories: cs.LG, stat.ML arXiv: https://arxiv.org/abs/2605.22765v1 Discrete diffusion models are often trained through clean-data prediction, but the prediction can be used in different ways to define the reverse dynamics. In Masked Diffusion Models (MDM) these choices largely coincide, whereas in Uniform Diffusion Models (UDM) they do not. We show that the standard plug-in bridge parameterization for UDM is not optimized by the denoising posterior, but by a leave-one-out posterior that predicts each clean token without using its own noisy observation. This identifies a mismatch between the plug-in ELBO and the usual cross-entropy denoising objective. We characterize the leave-one-out target and derive exact conversions between the denoiser, the leave-one-out posterior, and the score. These conversions allow us to disentangle parameterization and training objective. Our results also lead to inference improvements without any additional training through an informed predictor-corrector sampler and improved temperature sampling based on the leave-one-out predictor. We further introduce an absorbing-state reformulation of uniform diffusion that preserves the UDM joint law while decomposing it into masked-diffusion-like sampling operations, with simpler denoising posteriors, carry-over unmasking, and a natural remasking mechanism. On language modeling, leave-one-out parameterizations consistently improve UDM generation, while the absorbing construction matches or surpasses masked diffusion. These results suggest that the empirical gap between masked and uniform diffusion is driven less by the choice of marginals themselves than by parameterization and sampling design. The code and models can be found at https://github.com/samsongourevitch/rev_udm. --- ### Advancing Mathematics Research with AI-Driven Formal Proof Search Authors: George Tsoukalas, Anton Kovsharov, Sergey Shirobokov, Anja Surina, Moritz Firsching, Gergely Bérczi, Francisco J. R. Ruiz, Arun Suggala, Adam Zsolt Wagner, Eric Wieser, Lei Yu, Aja Huang et al. Published: 2026-05-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.22763v2 Large language models (LLMs) increasingly excel at mathematical reasoning, but their unreliability limits their utility in mathematics research. A mitigation is using LLMs to generate formal proofs in languages like Lean. We perform the first large-scale evaluation of this method's ability to solve open problems. Our most capable agent autonomously resolved 9 of 353 open Erdős problems at the per-problem cost of a few hundred dollars, proved 44/492 OEIS conjectures, and is being deployed in combinatorics, optimization, graph theory, algebraic geometry, and quantum optics research. A basic agent alternating LLM-based generation with Lean-based verification replicated the Erdős successes but proved costlier on the hardest problems. These findings demonstrate the power of AI-aided formal proof search and shed light on the agent designs that enable it. --- ### Towards a General Intelligence and Interface for Wearable Health Data Authors: Girish Narayanswamy, Maxwell A. Xu, A. Ali Heydari, Samy Abdel-Ghaffar, Marius Guerard, Kara Vaillancourt, Zhihan Zhang, Jake Garrison, Levi Albuquerque, Dimitris Spathis, Hong Yu, Hamid Palangi et al. Published: 2026-05-21 Categories: cs.AI arXiv: https://arxiv.org/abs/2605.22759v3 While ubiquitous wearable sensors capture a wealth of behavioral and physiological information, effectively transforming these signals into personalized health insights is challenging. Specifically, converting low-level sensor data into representations capable of characterizing higher-level states is difficult due to high phenotypic diversity and variation in individual baseline health, physiology, and lifestyle factors. Moreover, collecting wearable data paired with health outcome annotations is laborious and expensive, and retrospective annotation remains practically unfeasible, contributing to a scarcity of data with high-quality labels. To overcome these limitations, we propose a foundation model for wearable health that is pretrained on more than one trillion minutes of unlabeled sensor signals drawn from a large cohort of five million participants. We demonstrate that the joint scaling of model capacity and pretraining data volume leads to systematic improvements in performance, as evaluated on a diverse set of 35 health prediction tasks, spanning cardiovascular, metabolic, sleep, and mental health, as well as lifestyle choices and demographic factors. We find that this population scale representation unlocks label-efficient few-shot learning and generative capabilities for robust daily metric estimation. To further leverage this learned representation, we deploy a classroom of LLM agents to autonomously search the space of downstream predictive heads built on the model embeddings, showing broad performance improvements that increase with LLM model capacity. Finally, we show how integrating these downstream predictors into a Personal Health Agent can support model responses that are more relevant, contextually aware, and safe, and we validate this via 1,860 ratings from a cohort of clinicians. --- ### Lumberjack: Better Differentially Private Random Forests through Heavy Hitter Detection in Trees Authors: Christian Janos Lebeda, David Erb, Tudor Cebere, Aurélien Bellet Published: 2026-05-21 Categories: cs.LG, cs.DS arXiv: https://arxiv.org/abs/2605.22756v1 Random forests are widely used in fields involving sensitive tabular data, but existing approaches to enforcing differential privacy (DP) typically degrade performance to the point of impracticality. In this paper, we introduce Lumberjack, a differentially private random forest algorithm that achieves substantially higher utility by constructing large random decision trees and then applying aggressive, privacy-preserving pruning to retain only sufficiently populated nodes. A key component of our approach is a novel $(\varepsilon,δ)$-DP heavy hitter detection algorithm for hierarchical data, whose error is $O_{\varepsilon,δ}(\sqrt{\log h})$ for trees of height $h$ and may be of independent interest. This favorable scaling enables the use of significantly deeper trees than in prior work, leading to improved expressiveness under privacy constraints. Our empirical evaluation on benchmark datasets shows that Lumberjack consistently outperforms prior DP random forest methods, establishing a new state of the art. In particular, our approach yields substantial improvements in the privacy-utility trade-off for practical privacy budgets. Our findings suggest that carefully designed DP random forests can close much of the utility gap, highlighting a promising and underexplored direction for future research. --- ### Spectral Tail Auxiliary Learning for AI-Generated Image Detection Authors: Xingyi Li, Jiahui Zhang, Yiheng Li, Yun Cao, Wenhao Wang Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22751v1 As generative image models evolve rapidly, the perceptual gap between generated and real images continues to narrow, making AI-generated image detection increasingly challenging. Many existing methods exploit frequency-domain cues for detection, typically described as frequency-domain artifacts or high-frequency discrepancies. However, the specific and recurring spectral regularities remain insufficiently understood and characterized. In this paper, we systematically analyze the one-dimensional radial log-power spectra of real and generated images. We find that generated images do not necessarily exhibit higher or lower energy across the entire spectrum or high-band range. Instead, their spectra deviate from the power-law decay and show an anomalous uplift in the ultra-high-frequency tail. We term this phenomenon spectral tail uplift. We further attribute this phenomenon to nonlinear harmonic accumulation in trained generative models, suggesting that it can serve as a structural cue across generative architectures. Based on this observation, we propose Spectral Tail Auxiliary Learning (STAL), a frequency-domain auxiliary supervision framework for generalizable AI-generated image detection. STAL transfers spectral-tail cues from a tail-aware frequency teacher to a spatial detector during training, while all frequency-domain modules are discarded at inference time. Consequently, STAL introduces no inference overhead. Extensive experiments on 9 public datasets show that STAL achieves strong generalization and stability across generators, data distributions, and real-world scenarios. --- ### Cyber-Physical Anomaly Detection in IoT-Enabled Smart Grids Using Machine Learning and Metaheuristic Feature Optimization Authors: Adis Alihodžić, Eva Tuba, Milan Tuba Published: 2026-05-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.22749v1 Modern smart grids rely on dense measurement infrastructures, communication links, and intelligent field devices. Although this improves supervision and control, it also increases vulnerability to cyber-physical disruptions. Operators must distinguish physical incidents, such as faults or line disturbances, from malicious actions, such as false data injection or unauthorized command execution. This chapter investigates this problem using the well-known MSU/ORNL Power System Attack Dataset. The proposed method combines machine learning with genetic-algorithm-based feature selection. The objective is twofold: to classify attack and natural events accurately, and to determine whether a reduced set of physically informative PMU/IED measurements can support reliable detection. Several baseline models are evaluated, including logistic regression, RBF-SVM, XGBoost, Random Forest, and Extra Trees. The results show that tree-based ensemble models are the most effective for the considered dataset, with Extra Trees providing the strongest full-feature baseline. After feature selection, the GA + Extra Trees model reduces the clean PMU feature space from 112 attributes to an average of 27.4 attributes over five runs, while increasing macro-F1 from 0.9118 to 0.9212 and ROC-AUC from 0.9791 to 0.9837. These results indicate that many synchronized electrical measurements are redundant. A compact subset of phasor-based features can still provide accurate and interpretable anomaly detection in smart grids. --- ### Superhuman Safe and Agile Racing through Multi-Agent Reinforcement Learning Authors: Ismail Geles, Leonard Bauersfeld, Markus Wulfmeier, Davide Scaramuzza Published: 2026-05-21 Categories: cs.RO, cs.AI, cs.LG, cs.MA arXiv: https://arxiv.org/abs/2605.22748v2 Autonomous systems have achieved superhuman performance in isolation or simulation, yet they remain brittle in shared, dynamic real-world spaces. This failure stems from the dominant single-agent paradigm for physical applications, where other actors are ignored or treated as environmental noise, preventing effective coordination. Here we show that multi-agent reinforcement learning provides the essential safety scaffolding required for real-world interaction. Using high-speed quadrotor racing as a high-stakes testbed, we train agents to navigate complex aerodynamic interactions and strategic maneuvering with a variable number of racers. Through league-based self-play, agents evolve sophisticated anticipatory behaviors, including proactive collision avoidance, overtaking, and handling multi-agent physical interactions, including aerodynamic downwash. Our agents outperform a champion-level human pilot in multi-player races at speeds exceeding 22 m/s, while simultaneously reducing collision rates by 50 % compared to state-of-the-art single-agent baselines. Crucially, training with diverse artificial agents enables zero-shot generalization to safer human interaction. These results suggest that the path to robust robotic co-existence lies not in isolated safety constraints, but in the rigorous demands of multi-agent interaction. Multimedia materials are available at: https://rpg.ifi.uzh.ch/marl --- ### Plug-in Losses for Evidential Deep Learning: A Simplified Framework for Uncertainty Estimation that Includes the Softmax Classifier Authors: Berk Hayta, Hannah Laus, Simon Mittermaier, Felix Krahmer Published: 2026-05-21 Categories: cs.LG, eess.AS, stat.ML arXiv: https://arxiv.org/abs/2605.22746v1 Real-world sensor-based learning systems require uncertainty estimation that is both reliable and computationally efficient. Evidential Deep Learning (EDL) provides single-pass uncertainty estimation by modeling the class probabilities via Dirichlet distributions, where the Dirichlet parameters are predicted by a learned neural network mapping. However, this approach can lead to computational challenges, as Dirichlet expected objectives are more complex than standard supervised learning losses, complicating their analysis and implementation. We address this issue by approximating the objective of the first-order empirical risk minimization problem induced by EDL with a plug-in loss evaluated at the Dirichlet mean and show that, under mild assumptions, the approximation error decays with growing evidence for a broad class of loss functions, including mean-squared error and cross-entropy loss. As a special case, our analysis provides justification for the use of softmax in the context of uncertainty estimation, since under a particular evidence-to-Dirichlet mapping, our framework includes the standard softmax classifier. We validate the proposed simplified objectives on the Google Speech Commands dataset and show that they achieve predictive accuracy and selective prediction performance comparable to classical EDL, while being simpler to implement using standard deep learning losses and training pipelines. To the best of our knowledge, this empirical analysis is the first to obtain coverage-accuracy trade-offs for speech recognition tasks through EDL. --- ### SeqLoRA: Bilevel Orthogonal Adaptation for Continual Multi-Concept Generation Authors: Javad Parsa, Enis Simsar, Amir Joudaki, Thomas Hofmann, André M. H. Teixeira Published: 2026-05-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.22743v1 Parameter-efficient fine-tuning enables fast personalization of text-to-image diffusion models, but composing multiple custom concepts remains challenging due to representation interference. Existing modular methods either rely on expensive post-hoc fusion or freeze adaptation subspaces, which limit expressiveness and concept fidelity. To address this trade-off, we propose Sequential regularized LoRA (SeqLoRA), a constrained continual learning framework that jointly optimizes both LoRA factors via bilevel optimization. Theoretically, we establish strong convergence guarantees for our algorithm and model the residual layer activations as a matrix sub-Gaussian process to derive high-probability bounds on catastrophic forgetting. We further prove that learning the LoRA basis from data minimizes residual interference energy more effectively than frozen-basis methods. Experiments on multi-concept image generation demonstrate that SeqLoRA improves identity preservation and scalability across up to 101 concepts, while avoiding costly fusion and reducing attribute interference in composed generations. --- ### Ternary Decision Trees with Locally-Adaptive Uncertainty Zones Authors: William Smits Published: 2026-05-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.22740v2 Decision trees assign identical confidence to instances near and far from each split threshold. We introduce ternary decision trees, which augment each split node with an uncertainty zone of half-width delta. A decision-theoretic framework characterises the optimal zone width delta* as the solution to a node-local cost-minimisation problem; four formal properties are established: accuracy decomposition, a sufficiency condition for decided accuracy improvement, an exact efficiency characterisation (eta = Dec-Acc minus Acc_u, the accuracy gap between decided and boundary-uncertain predictions), and asymptotic consistency of the margin method. Instances within the zone receive predictions by weighted blending of both child subtrees and are flagged as boundary-uncertain. We propose and evaluate five delta-estimation methods: quality-plateau (plateau width of the split criterion curve), class-overlap (empirical class-distribution overlap), gain-ratio (split quality relative to split entropy), node-bootstrap (threshold variance under node-level resampling), and margin (SVM-inspired distance to the nearest cross-class training example). All methods reuse statistics already computed during standard CART split finding, requiring no external noise specification. Evaluated across 71 of the 72 OpenML-CC18 datasets with 5-fold cross-validation, all five methods with probabilistic routing significantly outperform standard CART on decided accuracy (Wilcoxon signed-rank, p < 0.001). The margin method achieves the best efficiency (0.104 accuracy gain per unit flagging rate), wins on 42 of 72 datasets, and requires zero hyperparameters. Analysis on Breiman synthetic benchmarks confirms margin is self-calibrating on clean data. On mammography, node-bootstrap achieves +0.71% decided accuracy by flagging 10.8% of cases as boundary-uncertain. --- ### Proxy-Based Approximation of Shapley and Banzhaf Interactions Authors: Santo M. A. R. Thies, Hubert Baniecki, R. Teal Witter, Eyke Hüllermeier, Maximilian Muschalik, Fabian Fumagalli Published: 2026-05-21 Categories: cs.LG, cs.AI, stat.ML arXiv: https://arxiv.org/abs/2605.22738v2 Shapley and Banzhaf interactions capture the complex dynamics inherent in modern machine learning applications. However, current estimators for these higher-order interactions trade off between speed and accuracy. To overcome this limitation, we introduce ProxySHAP. ProxySHAP reconciles the high sample efficiency of tree-based proxy models with a principled path to consistency via residual correction. On a theoretical level, we derive a polynomial-time generalization of interventional TreeSHAP to compute exact interaction indices for tree ensembles, successfully bypassing exponential tree-depth dependencies in prior methods. Furthermore, we formally analyze the residual adjustment strategy, characterizing the specific conditions under which Maximum Sample Reuse (MSR) corrects proxy bias without its variance scaling exponentially with interaction size. Extensive benchmarking demonstrates that ProxySHAP sets a new state-of-the-art standard for approximation quality, including in large-scale applications with thousands of features. By achieving the lowest error in both small- and large-budget regimes, ProxySHAP significantly outperforms the prior best estimators ProxySPEX and KernelSHAP-IQ, while also delivering superior performance on downstream explainability tasks. --- ### The Distillation Game: Adaptive Attacks & Efficient Defenses Authors: Youssef Allouah, Mahdi Haghifam, Sanmi Koyejo, Reza Shokri Published: 2026-05-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.22737v3 Distillation attacks create a deployment trade-off for model providers: the same outputs that make a model more useful can also make it easier to imitate. We study this trade-off through a minimax game between a utility-constrained teacher and an adaptive student. Our framework yields tractable one-sided response rules: an adaptive evaluation rule in which the student reweights high-value examples, and a teacher-side defense template that suppresses outputs most useful for distillation. From a cheap proxy for example value, we derive Product-of-Experts (PoE), a simple forward-pass-only defense that combines the teacher with a proxy student during generation. Empirically, adaptive evaluation reveals a large passive--adaptive gap: on state-of-the-art defenses, adaptive students recover substantially more capability than passive evaluation suggests on GSM8K and MATH. Under this stronger evaluation, the apparent robustness gap between expensive defenses and PoE narrows considerably, while PoE remains substantially cheaper and preserves higher-quality reasoning traces. Overall, our results suggest that strong distillation remains difficult to stop, and that progress on antidistillation should be judged against adaptive students rather than passive ones. Our code is available at: https://github.com/ysfalh/distillation-game. --- ### Optimization over the intersection of manifolds Authors: Yan Yang, Bin Gao, Ya-xiang Yuan Published: 2026-05-21 Categories: math.OC, cs.LG, math.DG, math.NA arXiv: https://arxiv.org/abs/2605.22736v1 Optimization over the intersection of two manifolds arises in a broad range of applications, but is hindered by the coupled geometry of the feasible region. In this paper, we prove that the regularities -- clean intersection and intrinsic transversality -- are equivalent, which yields a tractable projection onto the tangent space of the intersection. Therefore, we propose a geometric method that employs a retraction on only one manifold and updates the iterate along two orthogonal directions. Specifically, the iterates stay on one manifold, and the two directions are responsible for asymptotically approaching the other manifold and decreasing the objective function, respectively. Under intrinsic transversality, we derive the convergence rate for both the feasibility and optimality measures, and show that every accumulation point is first-order stationary. Numerical experiments on problems stemming from sparse and low-rank optimization, including fitting spherical data, approximating hyperbolic embeddings on real data, and computing compressed modes, demonstrate the effectiveness of the proposed method. --- ### ChronoMedKG: A Temporally-Grounded Biomedical Knowledge Graph and Benchmark for Clinical Reasoning Authors: Md Shamim Ahmed, Farzaneh Firoozbakht, Lukas Galke Poech, Jan Baumbach, Richard Röttger Published: 2026-05-21 Categories: cs.CL arXiv: https://arxiv.org/abs/2605.22734v1 Biomedical knowledge graphs (KGs) treat disease associations as static facts, but temporal information is crucial for clinical reasoning, e.g., a symptom diagnostic of one disease at age 3 may imply a different disease at age 13. Existing KGs such as PrimeKG, Hetionet, and iKraph do not encode when a finding becomes clinically relevant over the course of a disease. This limits their usefulness for longitudinal clinical reasoning and retrieval augmentation. We introduce ChronoMedKG, a temporal biomedical knowledge graph that contains 460,497 evidence-linked triples (filtered from 13M raw extractions) covering 13,431 diseases. Each association is tied to temporal components like onset window or progression stage, which are backed by PMID-traceable evidence and a multi-signal credibility score. The graph is constructed through a disease-autonomous multi-agent pipeline in which multiple frontier LLMs independently extract knowledge from PubMed and PMC literature. Only those relations are kept that are supported by multi-model consensus, survive credibility filtering, as well as ontology alignment. ChronoMedKG scored 92.7% agreement against Orphadata and adds temporal grounding for 6,250 diseases absent from HPOA, Orphadata, and Phenopackets, including 1,657 Orphanet-coded rare diseases. We further introduce ChronoTQA, a benchmark of 3,341 questions across eight task types (six temporal plus two static controls), with a 12-question supplementary probe. Frontier LLMs lose roughly 30 points moving from static to temporal questions; ChronoMedKG retrieval rescues 47-65% of their long-tail failures, against 17-29% for HPOA-RAG. As such, ChronoMedKG provides a crucial temporal axis for retrieval-augmented clinical systems that was previously absent. --- ### HarnessAPI: A Skill-First Framework for Unified Streaming APIs and MCP Tools Authors: Edwin Jose Published: 2026-05-21 Categories: cs.AI, cs.SE arXiv: https://arxiv.org/abs/2605.22733v1 Every Python function deployed as an LLM tool must today exist in two forms: an HTTP endpoint for human-facing clients and CI pipelines, and an MCP tool registration for agent runtimes such as Claude and Cursor. These representations share business logic yet diverge in all the surrounding machinery (routing, validation, serialisation, streaming, and schema maintenance), and they drift apart as the underlying code evolves. We present HarnessAPI, a Python framework that eliminates this duplication by treating a typed skill folder as the single source of truth. From one handler.py plus Pydantic schemas, the framework automatically derives a streaming HTTP endpoint with Server-Sent Events, an interactive OpenAPI/Swagger UI, and a zero-configuration MCP tool, all served from a single process. Dual-mode content negotiation lets the same handler serve SSE-streaming and JSON-returning clients with no handler changes. A dynamic code-generation mechanism ensures Pydantic type annotations propagate correctly to FastMCP's inspection layer, resolving a technical limitation that prevents naive closure-based registration. Measured across six representative skills using cloc, HarnessAPI reduces framework-facing boilerplate by 74% compared with a manually maintained dual-stack implementation (FastAPI server + FastMCP server). HarnessAPI subclasses FastAPI, inheriting its full middleware, dependency-injection, and deployment ecosystem. It is available at https://github.com/edwinjosechittilappilly/harnessapi and on PyPI (pip install harnessapi) --- ### Beyond Acoustic Emotion Recognition: Multimodal Pathos Analysis in Political Speech Using LLM-Based and Acoustic Emotion Models Authors: Juergen Dietrich Published: 2026-05-21 Categories: cs.AI, cs.CL, cs.HC, cs.SD, eess.AS arXiv: https://arxiv.org/abs/2605.22732v1 We investigate whether acoustic emotion recognition models can serve as proxies for the Pathos dimension in political speech analysis, as operationalised by the TRUST multi-agent large language model (LLM) pipeline. Using a Bundestag plenary speech by Felix Banaszak (51 segments, 245 s) as a case study, we compare three analysis modalities: (1) emotion2vec_plus_large, an acoustic speech emotion recognition (SER) model whose continuous Arousal and Valence values are derived via post-hoc Russell Circumplex projection; (2) Gemini 2.5 Flash, an LLM analysing the full speech audio together with its transcript in an open-ended, context-aware fashion; and (3) TRUST-Pathos scores from a three-advocate LLM supervisor ensemble. Spearman rank correlations reveal that Gemini Valence correlates strongly with TRUST-Pathos (rho = +0.664, p < 0.001), whereas emotion2vec Valence does not (rho = +0.097, p = 0.499). We further demonstrate, via a systematic quality evaluation of the Berlin Database of Emotional Speech (EMO-DB) using Gemini in an open-ended annotation paradigm, that standard SER benchmark corpora suffer from acted speech, cultural bias, and category incompatibility. Our results suggest that LLM-based multimodal analysis captures semantically defined political emotion substantially better than acoustic models alone, while acoustic features remain informative for low-level Arousal estimation. Future work will extend this approach to video-based analysis incorporating facial expression and gaze. --- ### Post-Training is About States, Not Tokens: A State Distribution View of SFT, RL, and On-Policy Distillation Authors: Dong Nie Published: 2026-05-21 Categories: cs.LG, cs.AI arXiv: https://arxiv.org/abs/2605.22731v1 Large language model post-training methods such as supervised fine-tuning (SFT), reinforcement learning (RL), and distillation are often analyzed through their loss functions: maximum likelihood, policy gradients, forward KL, reverse KL, or related objective-level variants. We study a complementary factor: the state distribution on which supervision is applied. For an autoregressive policy, a state is a prompt plus generated prefix. SFT trains on fixed dataset states, while RL and on-policy distillation (OPD) train on states induced by the current learner. We formalize post-training as state-distribution shaping and run a controlled smallscale study using Qwen3-0.6B-Base on GSM8K, with TruthfulQA and MMLU as retention evaluations. Our results show three phenomena. First, a mild SFT run improves GSM8K with little forgetting, while a stress SFT run causes substantial retention loss. Second, OPD from a degraded SFT teacher surpasses that teacher on GSM8K, TruthfulQA, and MMLU, despite using the teacher as its only supervision source. Third, a lightweight on-policy RL run improves GSM8K while preserving retention. These results support a state-centric view of post-training: the source and locality of training states can be as important as the form of the supervision signal. --- ### Multiple Neural Operators Achieve Near-Optimal Rates for Multi-Task Learning Authors: Adrien Weihs, Hayden Schaeffer Published: 2026-05-21 Categories: cs.LG, math.NA, stat.ML arXiv: https://arxiv.org/abs/2605.22724v1 We study the approximation and statistical complexity of learning collections of operators in a shared multi-task setting, with a focus on the Multiple Neural Operators (MNO) architecture. For broad classes of Lipschitz multiple operator maps, we derive near-optimal upper bounds for approximation and statistical generalization. On the lower-bound side, we establish a curse of parametric complexity and prove corresponding minimax rates. Together, these results show that shared representations across tasks do not increase the overall cost: multi-task operator learning follows the same scaling laws as single operator learning. We also compare MNO with a multi-task extension of DeepONet based on concatenated task inputs and show that, from a worst-case approximation-complexity perspective, both architectures satisfy essentially the same asymptotic rates. --- ### The Value of Covariance Matching in Gaussian DDPMs and the Lanczos Sampler Authors: Md Sahil Akhtar, Aymane El Gadarri, Vivek F. Farias, Adam D. Jozefiak Published: 2026-05-21 Categories: cs.LG, cs.AI, cs.IT arXiv: https://arxiv.org/abs/2605.22723v1 A central error measure in Gaussian DDPMs is the path-space KL divergence between the exact reverse chain and the learned Gaussian reverse process. This quantity is especially relevant for procedures such as classifier guidance, which perturb the entire reverse trajectory rather than only the terminal sample. Prior analyses show that standard isotropic reverse covariances suffer an unavoidable $Ω(1/T)$ path-KL error as the number of denoising steps $T$ grows. We show that matching the full posterior covariance breaks this barrier, yielding an order-wise improvement that reduces the path KL to $O(1/T^2)$. To make full covariance matching practical, we introduce the Lanczos Gaussian sampler (LGS), a training-free, matrix-free method for sampling from the optimal reverse covariance using only covariance-vector products, which are available through Jacobian-vector products of the posterior mean. LGS avoids dense covariance storage and auxiliary covariance models. We prove that LGS approximation error decays exponentially in the number of Lanczos steps, where each Lanczos step requires a single Jacobian-vector product. Empirically, using only just three such steps improves sample quality over strong diagonal-covariance baselines, including OCM-DDPM, across standard image benchmarks. This identifies full covariance matching as both theoretically valuable and practically accessible for fast DDPM sampling. --- ### Can AI Make Conflicts Worse? An Alignment Failure in LLM Deployment Across Conflict Contexts Authors: Andrii Kryshtal Published: 2026-05-21 Categories: cs.AI, cs.HC arXiv: https://arxiv.org/abs/2605.22720v1 AI models are already deployed in societies affected by armed conflict, and journalists, humanitarian workers, governments and ordinary citizens rely on them for information or for their work processes. No established practice exists for checking whether their outputs can make those conflicts worse. We tested nine model configurations from four providers (OpenAI, Anthropic, DeepSeek, xAI) on 90 multi-turn scenarios designed to surface misaligned behaviour in conflict contexts: false equivalence between documented atrocities, denial of genocide, and failure to recognise ethnic slurs, among others. When such outputs feed into journalism, humanitarian reporting, or public debate, they can deepen divisions in fragile societies. Failure rates span 6\% to 47\% between the best and worst performing models, which makes model choice a safety question in its own right and when users pushed for ``balance'' in cases where international courts have already assigned responsibility, five of nine configurations failed 80 to 100 percent of the time. We release the first evaluation framework for this domain and propose adding it to alignment evaluation portfolios. --- ### Reading Task Failure Off the Activations: A Sparse-Feature Audit of GPT-2 Small on Indirect Object Identification Authors: Mahdi Nasermoghadasi Published: 2026-05-21 Categories: cs.LG arXiv: https://arxiv.org/abs/2605.22719v1 We report a small, reproducible audit of which sparse-autoencoder (SAE) features of GPT-2 small fire differently on failed versus successful trials of the Indirect Object Identification (IOI) task. On 300 prompts, GPT-2 small reaches 79.7% accuracy; 146 of the 24,576 features in the layer-8 residual-stream SAE release of Bloom (2024) clear a Holm-corrected significance threshold and 105 reach a large effect size (|Cohen's d| > 0.8). The strongest single correlate of failure -- feature 17,491, d=+2.93, Neuronpedia label 'cryptographic keys' -- is essentially silent except when the prompt's transferred object is 'the keys,' on which GPT-2 small fails 93.3% of the time vs. 7.5% on the other seven objects (Fisher exact p = 8.79 x 10^-33). We put this correlate through three controls that a mechanistic claim should pass. (i) A causal ablation: zeroing feature 17,491 in the residual stream across all token positions of the 45 keys prompts does not restore accuracy (6.7% -> 4.4%); the feature is a correlate, not a sufficient cause at this layer. (ii) A representation baseline: a logistic regression on the raw 768-dimensional residual stream reaches 5-fold ROC AUC = 0.929, matching the top-100 SAE features (0.927); the SAE basis adds interpretability, not predictive power. (iii) A seed-robustness check: across five random seeds the keys-subset failure rate stays in 75.0--93.3% (the behavioural effect is real), but feature 17,491 is the top-|d| feature in only 1 of 5 runs. The methodological contribution is therefore the audit pipeline (cheap, model-agnostic, surfaces named correlates) rather than any single feature found through it. We release the code, the 300-prompt corpus, the 300x24,576 activation matrix, the ablation and baseline scripts, and the figures. The full pipeline runs on a laptop (Apple M3 Max, no discrete GPU). --- ### WorldKV: Efficient World Memory with World Retrieval and Compression Authors: Jung Yi, Minjae Kim, Paul Hyunbin Cho, Wooseok Jang, Sangdoo Yun, Seungryong Kim Published: 2026-05-21 Categories: cs.CV arXiv: https://arxiv.org/abs/2605.22718v1 Autoregressive video diffusion models have enabled real-time, action-conditioned world generation. However, sustaining a persistent world, where revisiting a previously seen viewpoint yields consistent content, remains an open problem. Full KV-cache attention preserves this consistency but breaks real-time constraints: memory footprint and attention cost grow linearly with rollout length. Sliding window inference restores throughput but discards long-term consistency. We propose WorldKV, a training-free framework with two components: World Retrieval and World Compression. World Retrieval stores evicted KV-cache chunks in GPU/CPU memory and selectively retrieves scene-relevant chunks via camera/ action correspondence, inserting them back into the native attention window without re-encoding. World Compression prunes redundant tokens within each chunk via key-key similarity to an anchor frame, halving per-chunk storage to fit 2x more history under a fixed budget. On Matrix-Game-2.0 and LingBot- World-Fast, WorldKV matches or exceeds full-KV memory fidelity at roughly 2x the throughput, and is competitive with memory-trained baselines without any fine-tuning. Project Page: https://cvlab-kaist.github.io/WorldKV/ --- ### Live Music Diffusion Models: Efficient Fine-Tuning and Post-Training of Interactive Diffusion Music Generators Authors: Zachary Novack, Stephen Brade, Haven Kim, Hugo Flores García, Nithya Shikarpur, Chinmay Talegaonkar, Suwan Kim, Valerie K. Chen, Julian McAuley, Taylor Berg-Kirkpatrick, Cheng-Zhi Anna Huang Published: 2026-05-21 Categories: cs.SD, cs.AI, cs.LG, cs.MM arXiv: https://arxiv.org/abs/2605.22717v1 Interactive streaming music generation promises the use of generative models for live performance and co-creation that is impossible with offline models. However, SOTA models exist in the discrete-AR regime, requiring industrial levels of compute for both training and inference. In this work, we investigate whether audio diffusion models, with their wide support in the open-source community but non-streaming bidirectional nature, can be repurposed efficiently into interactive models accessible on consumer hardware. By taking a critical look at the modern pipeline for block-wise outpainting diffusion, we identify critical inefficiencies during inference that result in strictly worse computational efficiency than their discrete-AR counterparts. We propose Live Music Diffusion Models (LMDMs), a simple modification of the generative diffusion process that recovers, and then outperforms, the inference complexity of the discrete Live Music Models (LMMs) through block-wise KV Caching. Unlike LMMs, LMDMs further enable stable post-training alignment through our novel ARC-Forcing paradigm, reducing error accumulation without any explicit RL or reward models. We demonstrate the application of LMDMs in a number of creative domains, including text-conditioned generation, sketch-based music synthesis, and jamming. We finally show how LMDMs can be used as a generative instrument in a real artist-AI collaboration, utilizing LMDMs as a "generative delay" to transform musicians' improvisation live for variable timbral effects while running locally on a consumer gaming laptop. --- ### Parametric Modular Answer Set Programs Made Declarative Authors: Jorge Fandinno, Yuliya Lierler, Torsten Schaub Published: 2026-05-21 Categories: cs.AI, cs.LO arXiv: https://arxiv.org/abs/2605.22716v1 In this paper, we explore the concept of modularity in first-order answer set programming (ASP). We introduce a new formalism called parametric modular logic programs, which allows defining subprograms with parameters and intensionality statements. We demonstrate how this formalism can capture the semantics of clingo-programs with collective control, a feature that enables structuring and instantiating subprograms. We provide theoretical foundations for modular ASP, illustrate its usefulness, and connect to traditional non-modular ASP. --- ### AnyMo: Geometry-Aware Setup-Agnostic Modeling of Human Motion in the Wild Authors: Baiyu Chen, Zechen Li, Wilson Wongso, Lihuan Li, Xiachong Lin, Hao Xue, Benjamin Tag, Flora Salim Published: 2026-05-21 Categories: cs.CV, cs.AI, cs.CL, cs.HC arXiv: https://arxiv.org/abs/2605.22715v2 As wearable and mobile devices become increasingly embedded in daily life, they offer a practical way to continuously sense human motion in the wild. But inertial signals are highly dependent on the sensing setup, including body location, mounting position, sensor orientation, device hardware, and sampling protocol. This setup dependence makes it difficult to learn motion representations that transfer across devices and datasets, and limits the broader use of wearable IMUs beyond closed-set recognition. We introduce AnyMo, a geometry-aware framework for setup-agnostic human motion modeling. AnyMo uses physics-grounded IMU simulation over dense body-surface placements to generate diverse and plausible synthetic signals, pre-trains a graph encoder from paired synthetic placement views and masked partial observations, tokenizes multi-position IMU into full-body motion tokens, and aligns these tokens with an LLM for motion-language understanding. We evaluate AnyMo on three complementary tasks: zero-shot activity recognition across 14 unseen downstream datasets, cross-modal retrieval, and wearable IMU motion captioning, where it improves average Accuracy/F1/R@2 by 11.7\%/11.6\%/22.6\% on HAR, increases zero-shot IMU-to-text and text-to-IMU retrieval MRR by 15.9\% and 28.6\%, respectively, and improves zero-shot captioning BERT-F1 by 18.8\%. These results support AnyMo as a generalist model for wearable motion understanding in the wild. Project page: https://baiyuchen.com/project/AnyMo. --- ### AMEL: Accumulated Message Effects on LLM Judgments Authors: Sid-Ali Temkit Published: 2026-05-21 Categories: cs.AI, cs.CL, cs.LG arXiv: https://arxiv.org/abs/2605.22714v3 Large language models are routinely used as automated evaluators: to review code, moderate content, or score outputs, often with many items passing through one conversation. We ask whether the polarity of prior conversation history biases subsequent judgments, an effect we call the accumulated message effect on LLM judgments (AMEL). Across 84,088 API calls to 12 models from 5 providers (OpenAI, Anthropic, Google, DeepSeek, and four open-source models), we present identical test items in isolation or following histories saturated with predominantly positive or negative evaluations. Models shift toward the conversation's prevailing polarity (d = -0.17, p < 10^-53). The effect concentrates on items where the model is genuinely uncertain at baseline (d = -0.36 for high-entropy items, vs d = -0.15 when the baseline is deterministic). Bias does not grow with context length: 5 prior turns and 50 produce the same shift (Spearman |r| < 0.01; OLS slope p = 0.80). And there is a negativity asymmetry: paired per item, negative histories induce 1.52x more bias than positive (t = 13.03, p < 10^-36, n = 2,733). Scaling helps but does not solve it (Anthropic: Haiku -0.22 to Opus -0.17; OpenAI: Nano -0.34 to GPT-5.2 -0.17). Three follow-ups narrow the mechanism. The token probability distribution shifts continuously, not at a threshold. The negativity asymmetry has both token-level and semantic components, though attributing the balance is exploratory at our sample sizes. Position does not matter: five biased turns anywhere in a 50-turn history produce the same shift. The simplest fix for evaluation pipelines is a fresh context per item; when batching is unavoidable, balancing the history helps. ---