Research

In a new book, Professor Eric So explores how the technology affects our thinking skills.

arXiv:2610.19143v1 Announce Type: new Abstract: While Vision-Language-Action models have enabled impressive cross-task transfer in zero-shot settings, \emph{adaptation} methods are often necessary to achieve consistent performance on novel tasks. However, even though a wealth of methods for adaptation have been proposed, it is unclear how to choose the best method based on available trajectories. We close this gap by studying the impact of the adaptation method under a variety of common data scenarios: on- and off-policy data, offline vs online adaptation, and whether the data contains optimal or suboptimal trajectories. Our evaluation spans simulated tasks in the LIBERO suite and real tasks on a Franka platform. In the offline data regime, critic-based steering approaches outperform supe…

arXiv:2610.19141v1 Announce Type: new Abstract: Learning a collective policy for state-based Cartesian manipulation across diverse robotic embodiments presents a fundamental challenge due to embodiment-dependent transition dynamics. Collective learning aggregates multi-robot experience to discover shared task structure, yet naive aggregation can average incompatible action modes when different embodiments realize the same Cartesian command through different transition behavior. This leads to strategy ambiguity and performance collapse on dynamically outlying embodiments. To address this, we propose PACE, a two-stage collective learning framework that separates shared task-space intent from embodiment-dependent transition dynamics. The first stage employs a representation layer pre-trained…

arXiv:2610.19122v1 Announce Type: new Abstract: Deep space missions increasingly require spacecraft capable of autonomously operating in environments with communication delays and bandwidth limitations. A key enabler of such autonomy is the ability to run learning-based algorithms directly onboard. In this work, we present a flight deployment on a 6U CubeSat that demonstrates (1) onboard training and inference of a thermal segmentation model on an NVIDIA Jetson TX2 NX GPU in space and (2) closed-loop attitude control using the segmentation model on a real-time, thermal image feed to track the Moon. In addition to this demonstration, we implement ROS 2 Foxy as a data transport middleware and integrate an open-source file system that provides strong data integrity guarantees and self-healin…

arXiv:2610.19097v1 Announce Type: new Abstract: As the Arctic transitions toward a seasonally ice-free state, capturing the complex morphology of submerged sea ice is crucial for advancing models of ocean-ice-atmosphere heat exchange and monitoring under-ice habitats. Currently, sub-surface ice mapping is predominantly achieved through acoustic sensors such as multibeam sonar. While sonar provides robust large-scale geometry, it lacks spectral and high-resolution textural information required to record fine morphological features - a critical limitation in under-ice environments, where sparse visual features and degraded sensing conditions challenge the capture of detailed structural and optical surface properties. We present a benchmark for vision-based under-ice 3D reconstruction using…

arXiv:2610.19089v1 Announce Type: new Abstract: Achieving human-like dexterous manipulation remains a central challenge in robotics. Learning from human hand-object interaction (HOI) datasets offers a promising pathway to diverse dexterous skills, yet existing physics-based approaches train specialist policies for individual trajectories, thereby limiting skill reuse at scale. We introduce DexMimic, a generalist dexterous hand-object tracking policy jointly trained across diverse HOI trajectories. Efficient joint training requires allocating each rollout to both an undertrained phase and a starting state that supports continued tracking. DexMimic realizes this insight through Value-Guided Reset Allocation, which adapts the initialization distribution online to prioritize difficult phases…

arXiv:2610.19082v1 Announce Type: new Abstract: Robotic systems that must acquire information about a specific target often face severe occlusion from sur- rounding structures or from the target itself, such that a single camera viewpoint provides only partial information. Active perception must therefore select viewpoints that progressively reveal unobserved regions. However, conventional local next- best-view (NBV) methods typically select viewpoints according to their immediate utility, which can overlook views that provide limited information initially but enable more informative obser- vations from subsequent viewpoints. We formulated local target perception as a short-horizon sequential view-planning problem and proposed a Receding-Horizon Next-Best-View (RH-NBV) planner that evalua…

arXiv:2610.19067v1 Announce Type: new Abstract: A wearable assistive robot must continuously estimate terrain, locomotion mode, gait phase, and walking speed. These quantities are commonly addressed by separate, terrain-specific models, with datasets rarely collected under a common protocol. This paper presents a visuo-inertial framework that connects these estimates through a causal pipeline using a chest-mounted camera and four lower-limb inertial measurement units (IMUs). Anticipatory Terrain Perception (ATP) classifies incoming terrain with 96.3% accuracy and anticipates terrain transitions by 1.46 s on average. Terrain-Aware Locomotion Mode Identification (TALMI) achieves 93.7% and 98.8% accuracy for its level-ground and stairs experts respectively, with mean transition latencies bei…

arXiv:2610.19044v1 Announce Type: new Abstract: Tiny robots have limited computation and memory, yet visual navigation systems often maintain more state than a task requires. We ask whether persistent memory can instead be shaped by the task. We present \textbf{RingNav}, a \(10.6\)k--parameter visual heading estimator that constrains its memory to the circular manifold \(S^1\). RingNav learns only the turn between consecutive images, while heading is composed directly on the circle. On the TartanAir Visual SLAM Challenge, RingNav reduces mean absolute turn error by \(27\%\) and accumulated heading error over $4$ sec by \(15\%\) relative to a similarly sized learned-memory model. The structured update also gives exactly zero turn for identical frames and exact cancellation under sequence r…

arXiv:2610.18991v1 Announce Type: new Abstract: Humanoid policies trained in simulation can transfer qualitatively to hardware while failing to transfer the force interactions learned during training, compromising torque efficiency, load regulation, and robustness to impacts. This failure can arise from actuator sim-to-real gaps because forces depend on actuator torques. We introduce Contact-UAN, a residual world model that preserves a simulator's physics priors while correcting actuator dynamics to align simulated rollouts with recorded hardware rollouts. For humanoid motion with contact, this requires contact-consistent replay: if hardware playback in simulation begins with misaligned contact wrenches, the residual model learns to correct reset artifacts rather than actuator discrepanci…

arXiv:2610.18950v1 Announce Type: new Abstract: Humanoid loco-manipulation requires coordinated whole-body interaction with the physical world, yet collecting robot demonstrations at scale remains costly. Human demonstrations offer a scalable alternative, but existing egocentric human supervision often omits locomotion and posture, while embodiment differences and controller dynamics hinder accurate end-effector execution. We introduce HumanVerse, a 500-hour dataset of open-world human loco-manipulation pairing egocentric video with synchronized body and hand motion. Our human-to-humanoid pipeline aligns camera viewpoints and arm appearance, and refines motion references and controller inputs offline to reduce end-effector errors while preserving whole-body coordination. We validate the p…

arXiv:2610.19212v1 Announce Type: new Abstract: World models of visual information predict future image frames conditioned on given action sequences. This predicted scene evolution is driven by deterministic changes such as controlled motion on a robot platform, as well as stochastic changes such as rendering of manipulated objects or previously occluded regions. Our key insight is that the spatial location of these stochastic changes is itself learnable. We leverage this insight by jointly regressing both the deterministic next-state prediction and the spatial uncertainty map, then using the uncertainty map to guide a conditional diffusion process. We propose Decoupled Drift-Diffusion World Model (D3WM), which uses a transformer trained with a Gaussian negative-log-likelihood objective t…

arXiv:2610.19201v1 Announce Type: new Abstract: We present LidarVerse, a real-time unified LiDAR world model for sensor simulation across tasks and sensors. Existing LiDAR simulators largely follow one of two complementary paradigms: reconstruction-based methods preserve the geometry of a captured scene but require lengthy per-scene optimization, whereas generative methods amortize learning across data but lack the consistency needed for long-horizon rollouts. LidarVerse brings both paradigms into a single, unified conditional generative framework. A universal tokenizer maps LiDAR range sequences into a shared latent space, where a bidirectional teacher learns a common interface for conditioning on a LiDAR prefix, an HD-Map, a persistent spatial cache, and optionally, text. Varying the av…

arXiv:2610.19154v1 Announce Type: new Abstract: Field pea has gained interest in the Midwest U.S. and has the potential to benefit Iowa cropping systems. This study evaluated 17 accessions and 7 Canadian varieties for agronomic and seed yield performance. This study found that PI accession trials exhibit genetic diversity in phenology and canopy traits, whereas the Canadian elite varieties demonstrated great yield potential under Iowa conditions. The accession genotypes matured between 75 DAP (late June) and 95 DAP (late July) across six site-years. The top-yielding Canadian varieties averaged 1961 kg/ha across two site-years. To comprehensively assess field pea performance, this study extracted time-series Unmanned Aerial Vehicle(UAV)-derived traits, assessed UAV-derived traits, and eval…

arXiv:2610.19108v1 Announce Type: new Abstract: Multimodal Large Language Models (MLLMs) have been widely applied in embodied agents and robotics for vision tasks. However, MLLMs suffer from limitations in spatial reasoning and perception. Post-training techniques have shown promise for improving these capabilities; yet, there are limited data-driven surveys of such approaches. In this survey, we consider approaches to fine-tuning pretrained MLLMs for spatial reasoning tasks, ranging from supervised to reinforcement learning, with an emphasis on the datasets used. We investigate themes of dataset construction, data modality and annotation, and evaluation of the resulting models. Based on these themes, we categorize research into three groups: dataset design, dataset construction and evalu…

arXiv:2610.19102v1 Announce Type: new Abstract: Aerial and remote sensing video is increasingly evaluated like web video, by posing multiple-choice questions to multimodal large language models (MLLMs) over sampled frames and reporting accuracy. This paper argues that this recipe measures the wrong aspect, and supports the claim with a convergent meta-analysis of a broad set of published results, organized in four evidence tiers, in which a claim becomes a position only when collected evidence agrees. Parameter scale exhibits a median of +3.4 accuracy points per tenfold increase on aerial benchmarks, far below a human gap of at least 19 points, whereas acquisition conditions control perception, as 315 of 318 evaluations in 19 within-paper contrasts lose accuracy when conditions such as al…

arXiv:2610.19071v1 Announce Type: new Abstract: Monocular 3D object detection methods increasingly rely on depth-guided transformers, but they operate entirely in the perspective image plane where spatial relationships are distorted by foreshortening. We propose MonoMIP, a dual-decoder framework that adds complementary 3D geometric reasoning through Bird's-Eye View (BEV) features constructed via Multi-Height Inverse Perspective Mapping. We perform IPM at K elevation planes spanning ground to rooftop, producing an ordered stack of height-conditioned metric BEV features that reduces above-ground misprojection and exposes depth-dependent appearance variation across planes. A BEV decoder with multi-scale deformable attention aggregates this multi-height evidence per query, while a depth-guide…

arXiv:2610.19063v1 Announce Type: new Abstract: As video models enable longer and more realistic interactive experiences, maintaining a world consistent with its history becomes essential. This requires both preserving previously observed content and ensuring coherent world evolution over time. World memory is a critical capability to evaluate alongside visual fidelity and video duration. We introduce WorldMemBench, a benchmark for systematically evaluating world memory in video world models. Across 260 test cases, it assesses static memory through the retention of human identity and appearance, object appearance and geometry, and environmental content and geometry, and dynamic memory through event progression and state persistence. Using controlled camera departures and multi-revisit tra…

arXiv:2610.19052v1 Announce Type: new Abstract: Video-based remote physiological measurement (RPM) has broad applications, while Vision-Language Models (VLMs) provide a promising opportunity to enable end-to-end physiological monitoring for future human-aware intelligent systems. However, existing RPM methods rely on task-specific models and cannot be seamlessly integrated with the multimodal understanding and interaction capabilities of VLMs. We revisit RPM with VLMs and provide empirical evidence for two major bottlenecks: generic visual semantic representations fail to preserve fine-grained physiological cues; while discrete token prediction in autoregressive generation is inherently misaligned with continuous numerical measurement. To address these challenges, we propose VITAL (Visual…

arXiv:2610.19025v1 Announce Type: new Abstract: Modern image generation largely relies on specialized generative models that leave the process of image construction implicit in learned parameters. We explore an alternative: general-purpose coding agents that construct images through executable programs. Our study of zero-shot drawing on AgentDrawBench, a 200-prompt benchmark, reveals a distinctive capability profile. Agents excel at preserving structure and satisfying precise visual constraints, achieving competitive results in domains where compositional accuracy matters more than photorealistic appearance, while remaining limited in perceptual richness. Their visual expressiveness varies with the drawing interface, and a qualitative study on reconstruction shows how procedural skills ex…

arXiv:2610.19007v1 Announce Type: new Abstract: World models are a powerful means of reasoning and planning, but many of these action-conditioned predictive models are high dimensional and computationally expensive. These challenges limit the number of action candidates that can be evaluated by a planner. We show that this cost is avoidable for planners over short horizons, which only require local information with respect to the current state. For a fixed state, the frames reachable by varying a low-dimensional action (a camera movement or an end-effector command) lie on a low-dimensional manifold. We estimate the local shape of that manifold directly via a network trained to estimate how each image region moves per unit of each action axis supervised by data collected from samples by th…

arXiv:2610.19134v1 Announce Type: new Abstract: Feature distillation is usually expected to add to logit distillation. We find that this expectation does not hold for two common feature losses in compact multilingual automatic speech recognition (ASR), and analyze why through their gradients. The gradient of a projection-based loss is orthogonal to the task gradient, while that of Centered Kernel Alignment (CKA), a similarity-based loss, is weakly aligned with it. CKA lowers the word error rate (WER) on non-English languages, but at the expense of data-rich English. Three loss-level remedies (one trained, two probed) all fall short, and the CKA gradient overlaps with the subspace English uses in the shared encoder by 97.5% to 99.5% (about 81% by chance). Guided by this diagnosis, we propo…

arXiv:2610.19118v1 Announce Type: new Abstract: Emerging decision models such as JEV provide typed interfaces that map task instructions and observations directly to choices, judgments, and scores. As these models are increasingly used for automated decision-making, understanding their reliability under adversarial manipulation becomes critical. We introduce DecisionRobust, a benchmark for systematically evaluating the robustness of decision models across textual and visual inputs. Our benchmark covers practical decision-making scenarios and a diverse suite of attacks, including semantics-preserving text perturbations, image perturbations, and prompt injection in both modalities. We curate evaluation sets through external validity checks and cross-model candidate selection. We evaluate 30…

arXiv:2610.19113v1 Announce Type: new Abstract: In-context learning (ICL) allows language models to adapt their behavior from instructions and demonstrations in a prompt, but accuracy alone does not reveal how contextual information controls their predictions. We study the acquisition and trainability of ICL with a controlled diagnostic that separates copying a nearby demonstration from integrating information across the prompt. Across 17 models, about a third prefer the copied answer, a few are undecided, and the rest prefer the rule-consistent answer, which goes with higher accuracy and with control by the requested value rather than the closest conflicting example. Checkpoint analyses show that models begin prioritizing task-relevant prompt tokens over conflicting examples long before…

arXiv:2610.19095v1 Announce Type: new Abstract: Chatbot Arena ranks language models with the Bradley-Terry model, which gives each model one score and so assumes Luce's choice axiom. Choice theory predicts where this fails: near-copies are told apart more decisively than their scores imply. On 1.8 million Arena votes, votes between versions of one model are 1.37 times as decisive as Bradley-Terry predicts, between sizes of one generation 1.26 times and between models of one lab 1.12 times; a release a year later gives consistent multipliers, and the effect survives alternative tie treatments, prompt subsets and topics, and timing checks. How alike two models' answers are worded and formatted tracks the effect. A correction with four to six parameters predicts held-out votes better than Br…

arXiv:2610.19048v1 Announce Type: new Abstract: Record linkage and demographic inference from names are routine demands in social science. Most approaches to these problems require labeled training data that are difficult or costly to acquire. I show that pre-trained language models perform both tasks accurately without requiring task-specific training. For record linkage, I introduce EnsembleLink, an algorithm that uses pre-trained embeddings alongside string-based scorers to retrieve a short list of candidates and then identify the likeliest matches. EnsembleLink matches or exceeds the performance of most prior approaches while generalizing across scripts and formats that those methods are unable to handle. For ethnicity inference from names, I show that large language models (LLMs) ach…

arXiv:2610.19028v1 Announce Type: new Abstract: Science fiction is widely credited with unsettling gender. This article asks whether that holds at the linguistic level, in how readily gendered reference gives way to its counterpart, a neutral term, or other social category, and whether it separates the genre from general fiction. It does so by masking man/men, woman/women, and person/people across same-period science fiction and general fiction corpora and classifying RoBERTa's five likeliest substitutes by gender. Retention and displacement are permutation-tested over five resampling rounds, and the underlying sentences are read closely. Both corpora cross the gender binary in both directions, but unequally: where woman is masked, the model predicts a masculine term about twice as often…

arXiv:2610.19026v1 Announce Type: new Abstract: Sparse attention reduces the KV-cache traffic that dominates long-context LLM decoding, but practical methods largely operate at coarse granularity. Token-level selection offers finer control, yet presents two challenges: low-dimensional scoring may lose query--key structure needed for accurate ranking, while gathering scattered selected tokens can offset the resulting latency benefit. We introduce Joint Subspace Sparse Attention (JSSA) to address both challenges. For selection quality, JSSA learns a shared low-dimensional query--key subspace by minimizing token-score preservation error. We upper-bound the resulting non-convex objective with a tractable marginal objective that admits a closed-form spectral solution; direct refinement of the…

arXiv:2610.19021v1 Announce Type: new Abstract: Recent work has questioned whether large language models (LLMs) can be useful for temporal event prediction. Despite vast pretrained world knowledge, LLMs often underperform simpler specialized models. We ask whether this knowledge is truly inaccessible, or whether current methods fail to present event prediction in a form LLMs can use. We argue for the latter and show that LLMs become effective event forecasters when the interface is designed around two choices. The first grounds each event in a natural-language description, and the second separates event semantics from the prediction space to prevent semantic bias from label associations. To test this, we introduce DescribeEvent, a benchmark of ~150K events across seven datasets and six do…

arXiv:2610.19003v1 Announce Type: new Abstract: Abstractive summarization remains challenging across languages, where linguistic quality and factual consistency may diverge, particularly in low-resource settings. We present a standardized zero-shot benchmark of eight closed-source and open-weight LLMs on size-matched English and Bengali XL-Sum test sets. Our evaluation combines lexical and semantic metrics with reference-free factuality verification and multidimensional LLM-as-a-Judge assessment, providing a comprehensive analysis of both surface-level quality and factual reliability in multilingual summarization. Results reveal substantial cross-lingual variation: GPT-5.6 Luna leads several English lexical metrics but does not maintain the same advantage in Bengali, where LLaMA 3.3 70B a…

arXiv:2610.19033v1 Announce Type: new Abstract: Low-rank adaptation (LoRA) applies every rank-one component of its update to every token, and mixture-of-experts variants replace this with a learned router that activates a fixed number of components. We argue that the number of components a token needs is a detection question with a classical answer. A token's responses to the rank-one components of a trained LoRA consist of a few large values on a Gaussian noise floor whose width matches that of an untrained adapter, and the floor carries cross-talk rather than adaptation. We introduce SQUELCH, a router-free mixture of rank-one experts that (i) soft-thresholds each token's responses at a multiple of its own robust noise level, so the active rank varies per token; (ii) chooses that multipl…

arXiv:2610.19016v1 Announce Type: new Abstract: Regret matching (and its variants) and counterfactual regret minimization are central algorithms in practice for normal-form and extensive-form game solving. Yet these have historically lacked the stronger theoretical convergence guarantees available to optimistic and extra-gradient methods. Recent work introduced IREG-PRM+, a regret matching variant which does have strong theoretical convergence guarantees. However, it relies on a simultaneous, extra-gradient update. IREG-PRM+ is an instance of the more general IR-PRM framework, where empirically optimistic and alternating updates often have better performance. However, the prior extra-gradient analysis does not extend to these cases. We introduce a new approach that covers both settings. F…

arXiv:2610.19015v1 Announce Type: new Abstract: Real measurements are too limited to train an accurate indoor pathloss predictor that works across a wide range of buildings. Large-scale synthetic pretraining offers a way forward, but real building layouts are also scarce, and accurate ray-tracing simulators are slow, expensive, and often proprietary. We address both problems by introducing a generator that automatically creates diverse indoor layouts and a fast, open-source approximate ray tracer that computes radio maps for them. Using this pipeline, we generate 128M synthetic samples, pretrain ResNet-101 and Swin-T encoder-decoders, and fine-tune them on high-quality benchmark data. Despite simulator mismatch, our ResNet-101 model reduces error by 15% on average relative to the best kno…

arXiv:2610.19013v1 Announce Type: new Abstract: Current scaling law formulations suggest that increasing the number of unique samples in a dataset improves model performance, while repeated exposure to the same samples provides diminishing value. In this work, we examine this assumption for image self-distillation models. Through systematic experiments across diverse datasets, namely Maxar, SATLAS, Waymo, nuPlan, Ego4D, ImageNet, and YFCC15M, we demonstrate that the effects of sample repetition on downstream performance are non-trivial and dataset-dependent. We identify settings where repeatedly training on a smaller random subset can outperform training on a larger set of unique samples. This finding highlights a limitation of existing scaling laws. To address this gap, we propose a modi…

arXiv:2610.19011v1 Announce Type: new Abstract: The Practical Molecular Optimization (PMO) benchmark standardized evaluation in molecular optimization, but it is built on simple property-based oracles that do not capture structure-based drug design. As the field has moved to docking-based objectives, evaluation practice has become unstandardized, and different methods report results on different docking tasks, often without a shared protocol for fair comparison or sample-efficiency constraints. To address this, we introduce Practical Molecular Optimization for Docking (PMO-Dock), a benchmark and protocol for docking-based optimization consisting of 25 tasks covering hit generation, lead optimization, and a new specificity task requiring strong on-target binding while penalizing off-target…

arXiv:2610.18883v1 Announce Type: new Abstract: Token-selection methods for on-policy distillation (OPD) prioritize student-visited positions using diverse signals, including uncertainty, teacher-student disagreement, and mismatch direction. We instead derive token importance from the distillation objective itself: how much can each position's loss decrease under an equal local KL budget on the student's predictive distribution? The answer is the utility $s_t=g_t^\top F_t^\dagger g_t$, where $g_t$ is the logit gradient and $F_t$ is the categorical Fisher matrix. For reverse-KL it equals the student-weighted variance of the student-teacher log-probability ratio, and for forward-KL it equals the Pearson $\chi^2$ divergence, so the appropriate utility follows from the choice of objective. On…

arXiv:2610.18880v1 Announce Type: new Abstract: High-throughput perturbation screens increasingly use machine-learning models to predict the effects of gene combinations that were never measured. A difficult case arises when both genes in a test pair were seen individually during training, but their combination was not. In this setting, standard OOD methods often fail because neither gene appears unfamiliar on its own. We therefore compare each model prediction with a simple additive prediction obtained by summing the two observed single-gene effects. The difference between them, which we call additive disagreement (AD), measures how much extra interaction effect the model introduces for an unseen combination. Across eight prediction methods and three Perturb-seq datasets, specialised neu…

arXiv:2610.18879v1 Announce Type: new Abstract: Protein function annotation operates over an expanding label space, making class-incremental learning (CIL) a natural framework for incorporating newly discovered functions without full retraining. Yet existing CIL methods have been developed largely in vision and have not been systematically evaluated over frozen protein representations. We conduct a controlled benchmark of nine representative CIL methods across five sequence-, structure-, and sequence--structure-based protein encoders and two enzyme function datasets at different scales, while keeping the encoders frozen and evaluation settings consistent across representations. We find that no method transfers reliably across encoders, and that these failures persist across alternative ph…

arXiv:2610.18878v1 Announce Type: new Abstract: Post-training reasoning LLMs often requires making a tradeoff between costly supervision and compute inefficiency: supervised finetuning typically requires reasoning traces, which is data inefficient since in many applications one only has prompt-answer tuples and lacks the groundtruth chain-of-thought (CoT) sequence. On the other hand, reinforcement learning (with verifiable rewards, for example) can operate without CoT data but it is much more compute-inefficient since it requires multiple autoregressive completions per optimization step, none of which are guaranteed to give a non-zero policy gradient. In this work we explore a middleground, a self-distillation technique that uses answer-only data, treating CoT as latent variables. This re…

arXiv:2610.18877v1 Announce Type: new Abstract: Time-series language models typically align numerical signals with language using labels, captions, or question-answer pairs. Although such supervision encourages linguistic representations, it does not explicitly require them to preserve information about how the underlying time-series signal evolves. We introduce forecast grounding, a pretraining principle that requires language-aligned time-series representations to remain predictive of the future evolution before they are adapted to downstream tasks. We instantiate this principle in LOTUS, which divides each time-series channel into non-overlapping patches and maps them into the LLM embedding space. It then predicts the next patch from the LLM hidden states of the time-series tokens, enc…

arXiv:2610.14209v1 Announce Type: new Abstract: The most accurate agents executing user-provided tasks are driven by large, cloud-hosted models, which solve them by reading the private data of the user. Agents powered by models small enough to run on edge devices keep that data local, but yield far worse results. A third option is to have cloud models write a program that devices execute: the data stays local, but as the program is written before any data is seen, it fails whenever a step depends on interpreting content that appears only at run time. We explore an intermediate architecture, in which the cloud-written program defers bounded questions over user data to a small local model through an ordinary function call. On AppWorld this improves the success rate of a code-only agent in a…

arXiv:2610.14190v1 Announce Type: new Abstract: Explaining a vision--language model (VLM) prediction requires tracing training-example influence to the multimodal evidence that mediates it, yet existing attribution methods largely operate only at the sample level. Recent latent-mediated influence methods provide feature-level attribution in language models, but lack the multimodal and spatial structure needed to ground influence in VLMs. We introduce MOSAIC (Multimodal Orthogonalized Sparse Autoencoders for Influence Characterization). It learns a shared sparse image--text latent space whose concepts are 1) sparse within each modality, 2) supported on contiguous image regions, 3) aligned to the phrases that describe them, and 4) approximately decorrelated. It then applies EK-FAC to decomp…

arXiv:2610.14186v1 Announce Type: new Abstract: Could model-free reinforcement learning algorithms actually be learning a model of the environment? In this work, we apply counterfactual interventions which hold everything about a state fixed, besides a targeted change, to explore if the value function of a Proximal Policy Optimization agent forms a world model of the Crafter environment. Despite only being trained to optimize reward, we discover that the value function learns concepts such as the rules governing when a diamond is collectible, water and food having an impact on health, heuristics regarding inventory, and more. Finally, we discuss the safety implications of our visualizations and some shortcomings of current reinforcement learning agents that will prevent us from building t…

arXiv:2610.14180v1 Announce Type: new Abstract: Organisations that want language-model agents for security work face a choice between open-weight models, which they can run and audit themselves, and far more expensive closed frontier models. We study that choice on CTI-REALM (arXiv:2603.13517), a benchmark in which an agent turns a threat-intelligence report into a working detection by querying a live telemetry store. We evaluate twenty-six models, nineteen open-weight and seven closed, under a common harness, scorer and panel of LLM judges. Our main finding is that an open-weight model can match the strongest closed models at less than a tenth of their cost, provided it is given enough turns. The benchmark's published limit on agent messages is a confound: a task that exhausts the limit…

arXiv:2610.14149v1 Announce Type: new Abstract: VETNET represents an ecosystem of live benchmark datasets and evaluation frameworks designed to accelerate AI-driven innovation in animal science and veterinary medicine through standardized, high-quality datasets and multi-institutional collaboration across academic and non-academic sectors. This framework addresses the gap between AI advances in human medicine and the fragmented state of veterinary AI development.

arXiv:2610.14146v1 Announce Type: new Abstract: Long-horizon terminal benchmarks measure whether an agent can sustain progress through extended workflows with process rewards, not only whether its final state passes a binary check. We present Long-Horizon-Terminal-Bench 2.0 (LHTB 2.0), an extension of Long-Horizon-Terminal-Bench that preserves its 46 original tasks while correcting task and evaluation-infrastructure defects and extending 31 tasks. LHTB 2.0 contains 77 tasks across 13 domains and eight task types. The extended tasks are produced through a systematic reconstruction-task pipeline. We start from a working scientific or engineering artifact, then use a parameterized reference implementation to generate multiple scenarios, reveal six as behavioral evidence, hold out three for e…

arXiv:2610.14136v2 Announce Type: new Abstract: Iterative sampling in diffusion models can lead to error accumulation due to exposure bias. The denoiser is trained on real images interpolated with noise, but at inference its inputs are constructed from its own imperfect predictions from the earlier steps, something it has never seen at training time. This might be one reason diffusion models often struggle without additional help, such as CFG. We introduce a correction step for diffusion models which attempts to reduce this training-inference gap. The corrector takes the base model's prediction as input and is trained against the same ground-truth target. This makes correction an explicit learning task, with training inputs provided by the base model. During sampling, at each denoising st…

arXiv:2610.14118v1 Announce Type: new Abstract: The rapid advancement of Large Language Models (LLMs) has driven their deployment toward general-purpose harness environments, where LLM-based agents operate in sandbox workspaces and interact directly with file systems and shells. While numerous post-training pipelines exist for tool-use agents, their counterparts for harness-based agents remain scarce and are predominantly one-shot and teacher-dependent: a stronger external model synthesizes tasks and collects trajectories, the target agent is trained once, and evaluation results are consumed by human researchers rather than fed back into the loop. Such pipelines are inherently bounded by the capability of the external model and the scope of human prior, leaving no room for self-improvemen…

arXiv:2610.14115v1 Announce Type: new Abstract: Large language models have significantly lowered the barrier to writing code. Naturally, this has also led to the development of data science agents that promise automating the design and execution of data science from task formulation to model development. Existing evaluations of agentic data science (agentic DS) assess downstream predictive performance and code correctness, providing preliminary insight in agentic DS capabilities. We argue that these evaluations overlook the fundamental principles of data science: methodological data science- and domain knowledge. As such, only assessing downstream performance masks methodological failures such as data leakage and inappropriate validation, limiting the development of robust agentic DS syst…

arXiv:2610.14005v1 Announce Type: new Abstract: Under the accelerated pace of AI development, any benchmark whose difficulty is calibrated with humans or current AI systems becomes circular, or saturates soon. How can we then evaluate machines that may become more capable than the current machines and the people who write the tests? One answer to the question hinges on inverse problems, inductive inference questions for which difficulty depends on the complexity of the underlying mechanism, given the observations and the previous knowledge. Levin's Kt, a computable variant of Kolmogorov complexity, puts a theoretical upper bound to that difficulty, balancing the simplest representation encoding for the mechanism and the logarithm of the time steps to run that mechanism. Previous tests der…