cs.CL · cs.LGOct 21, 2026new
Zhennan Wan, Jianfei Chen
Reinforcement learning (RL) is an effective way to strengthen the long-context reasoning of large language models (LLMs), but it needs training data that are long, diverse, challenging, and verifiable. Human annotation of such data is expensive, naturally occurring data are too scarce to sustain RL, and existing synthesis methods depend on external corpora or stronger teacher models, which become harder to obtain as the trained models get stronger. Recently, RL from zero data, in which a model learns from tasks it proposes itself, has achieved excellent performance, but existing methods target short-context tasks. Therefore, we propose $\textbf{LongZero}$, a method that lets an LLM improve its own long-context reasoning without any external data. Instead of writing questions and answers over existing documents, LongZero first has the model synthesize the question and answer with its supporting evidence in a compact structured form, and only then separately expands the evidence into a long multi-document context. LongZero combines these self-written multi-hop reasoning tasks with self-written multi-round retrieval tasks and trains the model with GRPO. We evaluate LongZero on two LLMs across five long-context reasoning benchmarks and compare it with RL on data synthesized from external corpora. Experimental results show that LongZero can generalize to real tasks, and outperforms both baselines, suggesting a viable and promising path to improving long-context reasoning of LLMs.
cs.CVOct 21, 2026new
Adrien Ramanana Rahary, Yutong Bai, Nicolas Dufour et al.
Video world models predict future frames from past frames and actions. They serve as simulators in which agents train, plan or are evaluated, and that humans can control interactively. They generate in the latent space of a video autoencoder, whose compression sets their cost per frame and how long each action waits to take effect. Yet world models often reuse the autoencoders of text-to-video generators, built for offline generation with no agent waiting and, without their training data and code, not retrainable at another compression. We therefore ask how, where and how much a world model should compress. We build CADRE, an open family of causal diffusion representation autoencoders spanning multiple spatio-temporal compressions. Its members differ only in a small bottleneck on frozen features and share a diffusion decoder self-distilled to one or a few steps. We train the same world model on each autoencoder in two domains, manipulation and navigation, with two environments each. CADRE forms most of the latency-quality frontier, especially with a moving camera. Our study also leads to four findings. At strong compression, a diffusion decoder is sharper than perceptual regression. In robot manipulation, temporal compression is more effective in the autoencoder than in the world model. Manipulation and navigation call for opposite compressions: temporal compression costs little when the camera is still, and spatial compression when it moves. Surprisingly, with a still camera, coarser spatial tokens hurt motion far more than image quality. Code, models and a live demo are at https://kyutai.org/blog/2026-10-21-cadre
cs.CVOct 21, 2026new
Zhichao Zhai, Menglei Zhao, Guikun Chen et al.
Language models' abilities to interpret tasks and generate programs offer a promising basis for 3D spatial reasoning. Yet following spatial instructions in 3D scenes requires programs to construct geometric references that language leaves implicit. In this work, we introduce VoxProg, which formulates 3D object placement through voxel programming: a language model translates task interpretations into executable programs over local voxel views of scene surfaces. Because these views index surface geometry on regular grids with explicit coordinates and neighborhood relations, an evidence program can isolate task-relevant regions and boundaries directly from scene meshes, and a downstream pose program can measure the resulting geometry to compute a target $SE(3)$ pose. Visual and numerical feedback from execution lets the model inspect intermediate selections and candidate placements to iteratively revise its code. To evaluate open-ended placement across functional use, geometric fit, and contextual arrangements, we introduce the 71-task VoxProg-Bench benchmark along with an evaluation suite pairing perceptual Intent Alignment and 3D Physical Validity. Across VoxProg-Bench and external benchmarks, VoxProg consistently outperforms relation-based, keypoint-based, and mesh-based baselines.
cs.RO · cs.AIOct 21, 2026new
Yunhao Luo, Wentao He, Yinpei Dai et al.
Existing world models typically perceive a scene through current visual frame or a narrow history window. However, these observations may not reveal the physics properties that determine transition dynamics, such as an object's mass distribution, joint damping, or surface friction. Without calibrating these physics factors for world models, predictions tend to regress toward an empirical prior and induce severe drift in novel test scenes. To this end, we propose PhysiCalWorld, which calibrates action-conditioned world models to the instance-specific dynamics via in-context prompting, without fine-tuning or analytical physics modeling. Our method encodes an exploratory interaction clip collected within the test scene into a compact physics descriptor token, which grounds subsequent predictions in the scene's actual physical behavior. PhysiCalWorld is trained end-to-end with the latent prediction loss alone, without physics labels. At inference time, the descriptor token is extracted once and can be reused across rollouts with negligible compute overhead. We evaluate PhysiCalWorld across four environments covering diverse physics variations and show that it substantially improves rollout prediction accuracy and visual planning success rate. The learned descriptors also form a physically structured latent space and generalize beyond the physics ranges seen during training.
cs.GT · cs.CCOct 21, 2026new
Ashkan Soleymani, Gabriele Farina
We give an algorithm for computing an $\eps$-approximate fixed point of any $d$-dimensional nonexpansion with respect to the $\ell_\infty$ norm in time $\log(1/\eps)^{O(\log^2 d)}$. At its core, our method uses a boosting argument to geometrically shrink a search region. Among other consequences, our algorithm implies that simple stochastic games can be solved in quasipolynomial time.
cs.LG · math.PROct 21, 2026new
Imon Banerjee, Pierre Alquier
We prove a self-normalized Bernstein inequality for bounded functions of a regenerative process with independent, identically distributed marked cycles of finite mean length. The bound uses observed cycle lengths and retains explicit terms for truncation and the cycle ending after the observation horizon. Selecting a tuning parameter from a finite deterministic collection gives a version of Bernstein inequality where all terms are empirical but one, related to the tails of the regeneration times. We then study this term in many regimes: it turns out that it is of second-order in many cases of interest. This is important for applications in machine learning: we prove generalization error bounds for classification with non-i.i.d., regenerative data. These bounds can be used to define learning objectives: by minimizing the first-order term only in the bound, we define agnostic classifiers that still satisfy oracle inequalities. We also show that the bound on the average deviation provided by our results is minimax-optimal in the sample size and in the expected regeneration time. Finally, we combine the Bernstein inequality with a regenerative strong approximation to obtain a Gaussian approximation using an observable block variance estimator.
cs.LG · cond-mat.dis-nnOct 21, 2026new
Francesco Mori, Melanie Weber
The success of convolutional neural networks is often attributed to locality and weight sharing. However, how these two properties separately shape learning, and how their effects depend on the spatial correlations of natural images, remain poorly understood. We address these questions in a teacher-student model with a convolutional teacher acting on $P$ input positions with correlation $c$ between positions. We compare convolutional and locally connected students trained by online stochastic gradient descent, and derive ordinary differential equations that describe their dynamics exactly in the limit of large feature dimension $D$. For uncorrelated positions ($c = 0$), the two students follow the same learning trajectory, with the convolutional student learning exactly $P$ times faster. We also find that locality affects the sample complexity: a student that matches the kernel size of the teacher recovers it from $\mathcal{O}_{D}(D)$ samples, whereas $\mathcal{O}_{D}(D\log(D))$ samples are needed with a larger kernel. For $c > 0$, spatial correlations can reduce the advantage of weight sharing. The asymptotic error of the convolutional student grows with $c$, whereas that of the locally connected student is independent of $c$ and overall lower when evaluated at the same learning rate. When the learning rate is instead tuned to the training horizon, however, the locally connected student wins at early times and the convolutional student at late times, with a transition that we predict analytically.
cs.CL · cs.AIOct 21, 2026new
Shibo Jin, Zhiqiu Xu, Yiling Zou et al.
Through a tangle of real-world interactions, from synthetic training pipelines to data shared on the open web, frontier LLMs absorb what their predecessors and competitors create. How well does each model play its role in this ecosystem, as a creator and as a solver? We introduce Environment Duels, a dynamic agentic benchmark that evaluates both capabilities on interactive environments. Each model writes environments, each with a privileged hint, aiming for tasks that capable solvers fail without the hint and pass with it. Across nine frontier LLMs, solving and authoring prove to be distinct skills: the strongest solvers are not the strongest authors, while smaller models learn more from how others played and write the hardest environments and the most useful hints. Beyond the scores, the pairwise interactions form a microcosm of how models affect one another. Head-to-head outcomes are transitive and so yield a stable ranking; independent authors converge on the same designs; and in a reinforcement-learning study with Qwen3.8-27B, the learner's hint lift on an author's environments tracks how much training on those environments improves it on held-out reasoning and agentic benchmarks, whether the environments are its own or another model's. Each of these patterns has a counterpart in how models shape one another outside the duel. By staging these interactions under controlled conditions, Environment Duels offers a window onto how LLMs teach, challenge and learn from one another. Code is available at https://github.com/LLM-Duels/EnvDuels.
cs.DC · cs.AIOct 21, 2026new
Clément Burgelin, Rachid Guerraoui, Zicong Hong et al.
Consensus is a fundamental problem in distributed computing and at the core of state machine replication (SMR). Despite decades of active human research, discovering practical low-latency solutions to the problem is still open, as it requires combining protocol knowledge, correctness reasoning, implementation, and experimental evaluation. We report on how AI helped us discover A-greement, a new consensus protocol for geo-replicated state machines that significantly outperforms state-of-the-art solutions. Averaged across six deployments with 5 or 9 replicas, A-greement reduces the mean latency of a geo-distributed key-value store over the fastest baseline by 12.2% and 33.3% on write-heavy and read-heavy workloads, respectively, and matches the fastest baseline on read-only workloads. A-greement combines lease-authorized local reads, direct client learning for leader-ordered writes, and a fast fixed-quorum per-key write path, selected according to network topology and workload. At the heart of our discovery lies AI-greement, an agent harness for the "autonomous" discovery, implementation, experimentation, and verification of candidate protocols. AI-greement integrates (1) a concept-oriented protocol knowledge base, (2) a modular implementation skeleton, and (3) a Lean-based verification framework with reusable safety definitions, reference specifications, and proofs. AI-greement is of independent interest: its Lean formalization uncovers several previously unreported vulnerabilities in existing consensus protocols, and it provides guidelines on how domain-specific harnesses can turn general-purpose coding agents into effective tools for (distributed) computing research. Yet we argue that human consensus experts remain essential before, during, and after the search: agents reshape the researcher's role rather than end consensus research.
cs.CL · cs.AIOct 21, 2026new
Houde Qian, Zheng Jiang, Wenyi Li et al.
Quantitative research is an iterative loop: a researcher proposes a market hypothesis, implements it as an executable signal, tests it on historical data, and decides what to keep. Existing benchmarks score formulas, code, or individual trading decisions, leaving open whether large language models (LLMs) can complete this loop and produce strategies that work beyond the research period. We introduce Quant-Arena, a benchmark and public leaderboard for end-to-end quantitative research on real A-share market data. A model constructs features and stock-scoring code, receives feedback only on 2022--2023, and freezes one strategy for assessment on the sealed year 2024. A cash-account evaluator applies trading constraints and costs, including market impact estimated from a simulated order book. Scores are dominated by realized excess return and complemented by signal quality and the diversity of features used by the strategy. We evaluate nine frontier models, among which GPT-6 Astra scores highest; yet no model beats the equal-weight market in 2024. Quant-Arena is open to submissions from any model or agent harness.
math.ST · cs.ITOct 21, 2026new
Shayan Hundrieser, Tudor Manole
We analyze the problem of estimating univariate Gaussian mixture densities using the nonparametric maximum likelihood estimator (NPMLE) based on $n$ independent and identically distributed samples. We show that the unconstrained NPMLE achieves the minimax optimal rate $\log(n)/n$ in squared Hellinger distance, when the true mixing measure is sub-Gaussian. In addition, we show that when the true mixing measure has compact support, the NMPLE constrained to a compact interval also achieves the minimax optimal rate, which is slightly faster and of order $\log(n)/(n \log \log n)$. Our proof consists of sharply characterizing the local Hellinger bracketing entropy of the class of finite Gaussian mixtures. One of our key technical innovations is to sharply bound the $L^2$ bracketing entropy of balls of bounded-degree polynomials, which may be of independent interest. As an application, we establish the regret-optimality of $g$-modeling procedures based on the NPMLE in the Gaussian Empirical Bayes problem with bounded priors.
cs.LGOct 21, 2026new
Yunpeng Mei, Xiaowen Zhu, Chenyu Wang et al.
Flow matching trains a continuum of velocity regression problems but usually reports a loss averaged over interpolation time, obscuring which regions are limited by intrinsic ambiguity and which remain data-limited. We introduce the Velocity Ambiguity Profile (VAP), $\mathcal{A}_v(t)$, the time-local Bayes-risk floor of velocity prediction, and decompose time-resolved loss as $L_N(t)=\mathcal{A}_v(t)+R(t;N)$, where only the excess term $R(t;N)$ is reducible by more data or a different estimator. For $N$ training samples and a $K$-component mixture in the structured example, we prove two statistical benchmarks. In an intrinsic local-regression benchmark, $\mathcal{A}_v(t)$ enters the time-dependent noise factor of the standard nonparametric rate. In separated isotropic Gaussian mixtures, the Bayes velocity becomes locally affine, and a label-oracle residual-regression benchmark has parametric excess-risk scale $K\mathcal{A}_v(t)/N$, matched within the same separated residual Gaussian-location subproblem. Controlled Gaussian-mixture experiments validate this diagnostic through a dense transition sweep, calibrated oracle-to-blind and empirical-proxy audits, and a local controlled time-sampler stress test in which a frozen reducible-error proxy reduces reducible loss while cross-cell sensitivity delimits aggressive sampling. On CIFAR-100 DINOv2 feature flow, guided time sampling reduces uniformly evaluated velocity loss by $6.78\%$ and global feature Fréchet distance (FD) by $23.6\%$, while class-conditional kernel inception distance (KID) shows no improvement. VAP turns an averaged flow-matching objective into a time-resolved diagnostic for locating ambiguity-limited and data-limited regions.
cs.RO · cs.AIOct 21, 2026new
Sriyash Poddar, Yanda Bao, Jacob Krantz et al.
Sequential decision-making under partial observability is difficult because agents must explore to infer hidden, task-relevant information before acting effectively. Learning such information-gathering behaviors tabula rasa via online reinforcement learning is sample-inefficient, whereas offline imitation averages expert actions across hidden states and fails to learn exploration. In this work, we propose ASTEROID, a simple and effective student--teacher supervised learning framework that can learn test-time exploration. Although a privileged expert never needs to explore, we demonstrate that a history-conditioned student can learn to do so when trained with on-policy contexts: the student first collects on-policy context under partial observability, and the privileged expert provides action labels conditioned on that context. By distilling expert actions directly conditioned on the student's on-policy history, the student learns to systematically resolve partial observability, i.e., search while uncertain and exploit once the hidden state is revealed. This enables efficient test-time exploration (similar to Bayesian posterior sampling) without incurring the impractical training-data costs of reinforcement learning. Across a diverse range of eleven different environments, including discrete exploration benchmarks, robotic manipulation, navigation, and Procgen environments, ASTEROID achieves ~2x higher returns under fixed exploration budgets, uses ~100x fewer interactions than RL baselines, and outperforms frontier models at exploration. Finally, our method transfers policies trained in simulation to vision-denied real-world robot manipulation, exhibiting efficient exploration and increasing success by +40% over standard distillation. Overall, it provides a simple yet surprisingly effective framework to learn efficient test-time exploration.
cs.LG · cs.AIOct 21, 2026new
Sonja Topf, Ana Sanchez-Fernandez, Günter Klambauer
Despite recent advances, current AI systems remain highly susceptible to domain shifts, particularly in biomedical imaging, where images are usually acquired under substantially different experimental conditions. Although recent domain adaptation methods have made substantial progress, significant performance gaps under domain shift persist. To measure progress in this regard, we introduce BatchBench, a leaderboard that evaluates and compares the robustness of AI systems on biomedical images from unseen experimental batches and data sources. BatchBench is built on JUMP-CP, one of the largest publicly available biomedical imaging datasets. Generated using the Cell Painting assay, JUMP-CP exhibits pronounced batch effects and domain shifts, making it a challenging testbed for evaluating robustness to distributional changes.
cs.CLOct 21, 2026new
Nurul Labib Sayeedi, Md. Faiyaz Abdullah Sayeedi, Zabir Al Nazi et al.
Many NP-complete (NPC) problems have a polynomial-time (P) twin that differs in a single structural detail, such as 3- versus 2-colouring a graph. A model that reasons rather than recalls should notice which side of this boundary a problem falls on, yet existing benchmarks either fix the wording of each task or ignore its complexity class, so they cannot separate the two. We introduce P2NPCBench, a benchmark of 12 problem families in which every NPC problem is paired with a P twin posed over the same surface. To remove the textbook cues that enable recall, we build it with AgentGAN, an adversarial multi-agent framework in which a generator rewrites each statement until a panel of frozen code models fails, while an executable validity gate keeps every rewrite answer- and class-preserving. The result is 2,114 hardened statements with hand-verified reference human solvers. Evaluating 12 proprietary and open-source models, we find that every model scores lower on NPC statements than on their P twins (35.4% vs. 63.8% pass@1), and that 95.1% of model-family pairs fail more often on the NPC side. A tenfold time limit raises NPC accuracy only from 28.5% to 31.9%, so failures stem from wrong or crashing programs rather than slow ones. Hardening also narrows the class gap from 24.8 to 14.6 points, showing that part of the apparent gap reflects familiarity with textbook phrasing rather than complexity class.
cs.LG · cs.AIOct 21, 2026new
Meng Ding, Mingxi Lei, Jie Zhang et al.
Reinforcement Learning with Human Feedback (RLHF) has become a central paradigm for aligning Large Language Models with human values. However, the reliance on sensitive preference data necessitates rigorous differential privacy (DP) guarantees to prevent user information leakage. Existing private alignment methods typically assume dense reward parameterizations, resulting in estimation error bounds that scale polynomially with the ambient {model size $d$}. Such dependence is prohibitive in modern {overparameterized models} where $d$ is vast. In this work, we study \emph{Private Sparse RLHF} problem to address this limitation, assuming that human preferences (reward model) are governed by a subset of $s^* \ll d$ relevant features. We provide the first comprehensive theoretical framework for differentially private sparse reward estimation from pairwise preference feedback under the Bradley-Terry-Luce model, and introduce efficient algorithms for both Sample-DP and Label-DP settings. Theoretically, under the squared $\ell_2$ parameter estimation error, we establish upper bounds of $\widetilde{O}(s^*/n + {s^*}^2/n^2\varepsilon^2)$ for $(\varepsilon,δ)$-Sample-DP and $\widetilde{O}(s^*/n + s^*/n\varepsilon^2)$ for $\varepsilon$-Label-DP based on randomized response, where $n$ is the training data size. For the lower bounds, we show that the Sample-DP rate is nearly optimal, and that the guarantee for our Label-DP method cannot be further improved within the randomized-response framework. We also provide a general lower bound for the problem with Label-DP. Empirically, our framework demonstrates superior efficacy over dense baselines across sentiment generation and safety alignment benchmarks, offering a significantly more favorable privacy-utility trade-off.
cs.RO · cs.AIOct 21, 2026new
Likui Zhang, Zhe Zhu, Zhihao Zhan et al.
Vision-language-action (VLA) models have shown promise in robotic manipulation, but imitation learning (IL) alone often struggles to master complex long-horizon skills. Reinforcement learning (RL) offers an effective avenue for further improving VLA policies. However, existing VLA-RL methods typically suffer from delayed and biased credit assignment due to trajectory-level sparse rewards, while imbalanced subtask mastery creates bottlenecks that limit overall performance. To address these challenges, we propose AtomicRL, an RL post-training framework that leverages atomic task structures for reliable credit assignment and adaptive exploration. AtomicRL introduces Semantic-Aware Policy Optimization (SAPO), which uses semantic-aware gating to constrain TD residuals from failed atomic subtasks from propagating across task boundaries, mitigating credit contamination while preserving intra-subtask credit assignment and value bootstrapping. In addition, Bottleneck-Adaptive Target Exploration (BATE) dynamically coordinates task sampling and local exploration incentives based on the mastery of individual atomic subtasks, directing learning toward the remaining bottlenecks. Finally, we develop a Real2Sim2Real pipeline that bridges online RL in simulation with real-world deployment, facilitating practical transfer to physical robots. On LIBERO-LONG, AtomicRL achieves absolute performance gains of 3.6% and 4.0% over PPO across two VLA architectures. On CALVIN ABC-D, it improves full-task performance by up to 9.5% over PPO. Combined with the Real2Sim2Real pipeline, AtomicRL further improves real-world task performance by an average of 21.3%. These results demonstrate that AtomicRL effectively improves the learning efficiency and manipulation performance of VLA policies on long-horizon tasks.
cs.CVOct 21, 2026new
Ruotong Liao, Meikel Kokowski, Xiaoxiao Sun et al.
Today's multimodal assistants are typically prompted to interpret the visual evidence presented to them. To offload the cognitive burden of remembering across long horizons, wearable assistants should instead proactively acquire and maintain useful memory as the user's surroundings evolve. Doing so on the user's behalf requires identifying what matters, understanding the scene in real time, and initiating memory maintenance in the background while the user stays focused on their activities. We introduce ProactEgoMem, a benchmark that evaluates proactive memory in streaming egocentric video. Three tasks ground the evaluation in ambient environmental needs: Information Capture (IC) preserves transient yet dense visual information, State Tracking (ST) follows changing object states, and Location Tracking (LT) records object movements. The benchmark defines its protocol at two levels of proactiveness: the Notification Level asks for a single timely notification that satisfies the user's intent, while the deeper Intervention Level asks the agent to maintain memory on the user's behalf through an active start-to-stop tool lifecycle. Parallel ProactEgoMem-Daily and ProactEgoMem-Sci tracks contain 300/100 clips covering daily and scientific use cases, paired with 1000/500 questions in ProactEgoMem-QA, which measures the quality of the proactively acquired memory. Across proprietary, open-source, video-language, and streaming baselines, even the strongest models fail to capture memory both timely and accurately. ProactEgoMem provides a rigorous evaluation foundation for proactive multimodal assistance and for video agents that co-exist with their users.
cs.LGOct 21, 2026new
Ivan Žugec, Joshua Rosenthal, Ivor Lončarić et al.
Generative models propose crystalline materials by learning composition and geometry, but unconditional sampling gives no direct control over the space group, which constrains whether responses such as piezoelectricity are symmetry-allowed. We present Orbitol, a crystal generator built on the Crystalite diffusion Transformer that replaces atom tokens with Wyckoff-orbit tokens. Given a space group and a collection of occupied Wyckoff positions, Orbitol generates the atom type and free coordinates of each orbit together with a compatible lattice, so every valid structure is invariant under the operations of the requested group. It generates structures two to three orders of magnitude faster than the symmetry-aware baselines at the evaluated settings. On MP-20, 13.8% of Orbitol's attempts yielded stable or metastable, unique and novel structures that retained the requested space group after relaxation, compared with 10.8% for SGFM, the next best model. We also introduce SPARC, a score that combines retention of the raw detected space group with agreement between the retained-group and training distributions. Finally, we used Orbitol to generate candidates in piezoelectric crystal classes. Of the structures selected by internal screening, 89% passed our density functional theory checks, and 79% also matched no Materials Project or Alexandria entry with the same space group. Their median figure of merit was nine times that of the piezoelectric materials computed in the Materials Project, and 24% of the new candidates ranked in that set's top 10%.
cs.LGOct 21, 2026new
Hadi Khalaf, Sajani Vithana, Kunal Talreja et al.
Language models make it easy to build classifiers by simply prompting a model with the sample to classify. However, their predictions are neither deterministic nor uniformly reliable across samples. How can we combine predictions from different models to improve reliability? We view LLM classifiers as noisy measurement instruments of an underlying target label. Our goal is to maximize average classification accuracy at a given average query cost. To this end, we introduce circuits that organize LLM classifiers into an adaptive sequence of queries and decisions. We then characterize the optimal circuit, obtain the fundamental limits of achievable accuracy, and provide practical insights on realizing the optimal circuit. Across seven classification tasks spanning safety, health, political science, human preference, agent evaluation, decision-making and reasoning, our method improves the cost-accuracy frontier over routing and cascading baselines.
cs.LGOct 21, 2026new
Amirardalan Dehghanpour, Arian Noori, Ehsan Tanghatari et al.
Personalized federated learning must decide which clients should share representations. We introduce HDC-FL (Head-Derived Clustering for Personalized Federated Learning), which reads that decision off a signal shared-encoder training already produces: the client-specific classifi- er heads. At periodic refresh rounds the server projects the cached heads into a normalized low-rank descriptor space, groups the descriptors, builds one encoder prototype per group from the cached client encoders, and returns to each client a soft mixture of those prototypes paired with its own unchanged head. Client-specific heads are never averaged or distributed to oth- er clients; the shared-encoder trajectory is left intact; and no second optimization trajectory, additional local objective, or post-training fine-tuning stage is required. Across 11 CIFAR-10 and CIFAR-100 configurations covering label-subset, Dirichlet and rotation heterogeneity with ResNet-18 and MobileNetV2 encoders, HDC-FL is competitive with the strongest personalized and clustered baselines in macro-averaged mean accuracy and attains the numerically highest macro-averaged worst-client and bottom-20% accuracy among the evaluated methods. The tail results distinguish methods that look similar when only their means are compared. A controlled donor-swap diagnostic further finds positive donor-ranking associations between head similarity and encoder compatibility in all three settings tested, most clearly under covariate shift.
cs.MA · cs.AIOct 21, 2026new
Hanmin Li, Yury Demidovich, Peter Richtárik
Parallel AI agents can produce useful local edits yet degrade a shared solution when their edits are combined. We model this coordination problem as discrete block optimization with an exactly evaluable objective and a known dependency graph. Independent sets certify compatible edits whose verified gains, measured against the same current solution, add exactly. A coordinator constructs a safe compatible candidate and can also evaluate interacting edits to capture favorable interactions. We prove that an optimal fractional coloring yields a merger that preserves a $1/χ_f(G)$ fraction of the total verified local improvement, where $χ_f(G)$ is the graph's fractional chromatic number. This guarantee is tight among mergers that select a subset of the same proposals. Under conditional local-improvement guarantees and a block error-bound condition, repeated collaboration converges geometrically in the objective gap. The rate separates task structure, agent allocation, and computation cost, yielding a linear program for agent assignment and an explicit time-cost trade-off. Code-repair experiments show that verified merging resolves conflicts between individually useful edits that cause naive merging to fail. On larger conflict-free tasks, parallel agents reduce modeled completion time at higher input-token-equivalent cost. On three public road networks, fractional integration retains 922 units of verified graph-cut improvement across 90 paired runs. For route-search agents sharing exclusive access to vehicle routes, fractional scheduling reduces measured wall time by a median of 15% across twelve paired runs, with route-cost changes varying across runs.
cs.AIOct 21, 2026new
Xinzhe Wang, Xudong Weng, Zang Li et al.
Data agents built on ReAct-style loops now execute complex analytical workflows over structured data, yet their intermediate reasoning still runs almost entirely through text. Human analysts instead move between numbers, tables, and charts, reading visual evidence to expose trends and anomalies that text summaries drop. We ask whether charts can serve as internal epistemic actions inside a data agent's reasoning loop, and how to teach an agent to use them. We first build \probe{}, a suite of 200 controlled tasks under four matched interface conditions, and find a persistent gap between what agents can read and what they choose to use. Agents score as well from self-generated charts as from numeric output, yet plotting access alone does not translate into consistent use. Task slices show that charts lead on temporal, pattern, and anomaly reasoning, while text leads on precise reading and conditional verification. We then propose \method{}, which distils two families of experience from conditioned trajectory rollouts. One family records how to select and verify analytical operations, and the other records when to invoke text, charts, or both. A trajectory judge scores outcomes and labels intermediate steps, and a validation gate admits only experience that improves held-out performance. On three held-out benchmarks with four LLM backbones, \method{} raises analytical accuracy and visual evidence quality over matched baselines, and backbones that never drew charts begin using them, all without touching model parameters.
cs.LGOct 21, 2026new
Adrien Deschenaux, Nadezhda Ilieva, Justin Deschenaux et al.
Recent Diffusion Language Models (DLMs) use Self-Conditioning (SC), or Loopholing, to improve their performance. While SC is traditionally seen as an auxiliary heuristic, its empirical success raises the question: can a continuous generative model be based purely on Self-Conditioning? We answer affirmatively with LAW (``Loopholing All the Way''), which propagates the Expected Embedding under the denoiser's prediction across steps, decoding the tokens once, at the end. On Sudoku, LAW outperforms previous Discrete and Continuous Diffusion Models, including Loopholing Discrete Diffusion. On Maze, LAW is the most accurate from 2 to 64 NFE. After training on TinyGSM and sampling at temperature 1 with 1024 NFE, LAW solves 42.2% of GSM8K problems against 34.5% for uniform Loopholing Discrete Diffusion. Additionally, fine-tuning a pre-trained Uniform Diffusion Model (Duo) with LAW increases its accuracy from 17.2% to 48.1%. Finally, on OpenWebText, LAW pushes the Pareto frontier and approaches Autoregressive Models (ARMs) with 64 NFE. Find our code at https://github.com/ilievanadezhda/law.
cs.CLOct 21, 2026new
Quang Anh Kiet Tran, Nam Anh Tran, Manh Hung Trinh et al.
Vietnamese pre-trained encoders have improved rapidly, but the strongest ones are large, multilingual-vocabulary models or depend on external word segmentation, which makes them costly to deploy. We present GoiBERT, a family of two compact Vietnamese encoders with 26M and 49M parameters that operate directly on raw text. GoiBERT learns from two complementary teachers. A same-tokenizer Vietnamese encoder first teaches general-purpose representations. A much larger teacher then provides task-level guidance through soft labels that are generated without ever labeling its own training data. Across eight benchmarks covering text classification, aspect-based sentiment analysis, natural language inference, and named entity recognition, GoiBERT-large achieves the highest average score among non-teacher models. It does so with fewer than half the parameters of the strongest compact baseline and twice its throughput. GoiBERT-base retains 99% of that baseline's performance at a quarter of its size. Our analysis shows that compression is nearly lossless for sentiment and classification tasks but remains costly for inference and entity recognition, which points to clear directions for future work. Our code and models are available at: https://github.com/Nhantran65/GoiBERT.git
cs.CL · cs.LGOct 21, 2026new
Saiteja Utpala, Chieh-Hsin Lai, Yuki Mitsufuji et al.
Uniform diffusion language models (UDLMs) generate text by iteratively denoising sequences of ordinary vocabulary tokens, any of which may be corrupted and subsequently revised. During training, UDLMs must denoise from a context that may itself be unreliable, whereas masked diffusion language models (MDLMs) always condition on a clean context. This challenging training objective makes UDLMs more FLOP-intensive to train than MDLMs, but it yields stronger downstream reasoning performance. Recently, selective looping, which recursively applies a few early Transformer layers within each denoising step, has been shown to improve training efficiency in MDLMs under an iso-FLOP budget. We investigate looping in UDLMs while keeping the training objective and sampler unchanged. We find that looping significantly improves training efficiency: looped UDLMs reach the validation NLL of a non-looped UDLM trained on about 0.5 trillion tokens with up to 2.8x fewer training FLOPs on OpenWebText and 3.3x fewer on FineWeb-Edu. We also find that, at a matched NFE budget, looped UDLMs reach up to 9.8% lower generation perplexity than the non-looped UDLM, with most gains achieved at moderate loop counts. Finally, we study different loop-count distributions and find that sampling smaller loop counts more frequently during training improves inference efficiency: on downstream tasks, this strategy uses 12% fewer loops than uniform loop-count sampling without degrading accuracy.
cs.RO · cs.AIOct 21, 2026new
May Espinola, Qifan Zhang, Yu Xiang
Recent advances in frontier AI models have demonstrated remarkable capabilities in visual understanding, reasoning, planning, and tool use, raising the possibility of using general-purpose AI models for robot manipulation without task-specific training. However, their manipulation capabilities have not been systematically evaluated on standardized robotics benchmarks. In this report, we present a unified closed-loop manipulation framework for evaluating frontier AI models on the LIBERO-10 benchmark. Our framework enables models to control a robot directly through visual observations, robot proprioception, and structured tool calls, primarily by specifying target end-effector poses and gripper commands, without task-specific fine-tuning or pretrained manipulation skills. We evaluate six frontier AI models, including GPT-6 Astra, GPT-6.1 Sol, GPT-6 Luna, Claude Opus 5, Claude Opus 5.5, and Gemini-4 Argon, across three reasoning effort levels, resulting in 18 configurations and 1,800 evaluation episodes. Our results show that GPT-6 Astra achieves the highest success rate of 44%, followed by Gemini-4 Argon at 43%. Increasing reasoning effort generally improves manipulation performance, but the benefits vary across models and come at additional inference cost. We also observe substantial differences across tasks, with models performing relatively well on object placement tasks but struggling with precise spatial arrangements and complex multi-step interactions. These findings demonstrate both the potential and limitations of frontier AI models for robot manipulation, highlighting the challenges of translating reasoning capabilities into reliable physical actions. We release our evaluation framework and experimental results to support reproducible benchmarking and future research on reasoning-based robot manipulation.
cs.SE · cs.AIOct 21, 2026new
Lorenz Brehme, Benjamin Steenhoek, Ruth Breu
Coding agents are typically evaluated using one-shot, well-formulated prompts. Benchmarks such as SWE-bench use existing GitHub issues as input for a coding agent and evaluate whether the agent successfully solves the task. However, in real-world use cases, prompts are often not as well-formulated or structured as existing GitHub issues. Additionally, a coding agent session is not limited to a single prompt; it typically involves a multi-turn interaction in which the user iteratively adds new tasks and requirements that the agent must fulfill.
To benchmark these abilities, we developed SWE-DIVE, a dynamic, interactive validation and evaluation framework for benchmarking coding agents in a multi-turn environment. To achieve this, we introduce a user simulator that can simulate different behavioral patterns and communicate with the coding agent. We also introduce a conversation validator that evaluates the user simulator's outputs to determine whether they fulfill the initial requirements and guide the agent toward successfully solving the task. The entire framework is embedded in a Harbor environment, which orchestrates the coding agent, user simulator, and conversation validator, and executes the test environment to determine whether the agent successfully solves the task. Our results and source code are available in our GitHub repository (https://github.com/lorenzbrehme/swe-dive_usersim).
cs.AIOct 21, 2026new
Tianyou Lai, Wentao Yue, Zhiwei Zuo et al.
Decision models map stated rules and observed states to discrete answers. Aggregate accuracy leaves unresolved which rule and information changes guide those answers. We connect controlled input changes to transfer, joint reliability, and directed information use through executable arithmetic rules. Semantic rule separation and crossed rule and state tests make transfer and joint success separately measurable. We compare equivalent wordings, calibrate joint correctness through marginal-preserving reassociation, and alter tables to define alternative answer targets. Adaptation gains increase with nearest-source answer agreement; fully conflicting quartets yield small pooled gains and different gain signs across source splits. Equivalent rewrites have opposing task-level effects, linking transfer gains to rule presentation. Joint-score departures depend on which rule and state associations the reference preserves. A decision service changes answers more across equivalent wordings than across identical-text repeats and usually follows the answer targets implied by altered table values or mappings.
cs.RO · cs.AIOct 21, 2026new
Zhaohui Yang, Quanbao Lin, Chengqi Shi et al.
Humanoid loco-manipulation requires an effective interface between high-level task policies and whole-body control. Existing interfaces typically rely on a single action representation for both task intent and whole-body coordination. This may burden the high-level policy or leave one aspect underspecified. We present SP-HLM, a framework built on a sparse-pose-driven interface. The interface combines five task-critical body poses with a learned latent motion prior to represent spatial intent and whole-body coordination in complementary forms. SP-VLA, a vision-language-action model, first predicts task-relevant sparse poses and finger joint positions from multimodal observations and language instructions using asynchronous action-chunk inference. Given these sparse targets, SP-MGM, a motion generation model trained through two-stage distillation, reconstructs coordinated full-body motion in real time. A pretrained motion encoder maps this motion to a compact latent representation. SP-GMT, a general motion tracker, then integrates the explicit sparse targets with this implicit motion prior for whole-body control. Experiments in simulation and on a real humanoid robot demonstrate improved motion tracking, whole-body coordination, and loco-manipulation success over existing approaches. Additional results and videos are available on our \href{https://sp-hlm.github.io/}{\textcolor{blue}{project page}}.
cs.AI · cs.LGOct 21, 2026new
Sudipta Paul, Pengrui Quan, Aounon Kumar et al.
Harness optimization, which improves an agent's prompts, tools, and workflows, is an increasingly effective way to enhance agents built on frozen or closed-source LLMs. Leading approaches such as the Darwin--Godel Machine and the Huxley--Godel Machine (HGM) rely on recursive self-improvement and were developed mainly for coding, where the competence needed to improve the agent coincides with the competence being improved. Planning tasks, in which an agent must produce a sequence of tool calls that jointly satisfies interacting constraints, break this alignment. There, open-ended self-modification repeatedly explores ineffective changes and consumes substantial evaluation budgets. We introduce HGM-Dual, a diagnostic-guided framework for budget-efficient harness evolution for planning agents. First, HGM-Dual represents agent evolution as lineage-aware context. This context records what each ancestor changed, which behaviors improved or regressed, and which modifications were ineffective, giving the editor explicit edit-outcome attribution instead of an unstructured archive of past agents. Second, HGM-Dual introduces error-conditioned dual-path child generation. For each selected parent, it generates a general child that preserves open-ended exploration and specialized children that target the harness components most likely responsible for the parent's highest-ranked constraint violations. HGM-Dual outperforms the strongest baseline by up to 14.45% relative improvement in case accuracy while matching HGM variants' final performance with 2x-4x fewer evaluation budget.
cs.AIOct 21, 2026new
Zijian Jiang, Xilu Wang
Text skills can extend the capabilities of tool-using agents, but long skill descriptions increase computational costs. Soft prompts provide a compact alternative, yet learning from entire successful trajectories may also introduce unnecessary behavioral changes. We find that the useful effect of a text skill is highly concentrated in a small subset of trajectory positions. Based on this observation, we propose Behavioral Localization, which compares the frozen agent with and without the text skill to identify positions where the skill both increases the likelihood of successful tokens and meaningfully changes the next token distribution. We use these localized positions to supervise a short soft prompt. Experiments on SpreadsheetBench with a frozen Qwen3.6-35B-A3B agent show that the top 10\% of trajectory positions capture 96.12\% of the total positive skill gain. In the shared prompt setting, selective supervision improves success from 30.36\% with full trajectory to 36.79\%. Under task-specific conditions, using the top 10\% localized positions achieves 57.38\%, outperforming both the full text skill and the on-policy OPCD baseline. These results show that a small set of behaviorally relevant trajectory positions can provide effective supervision of soft skills.
cs.CVOct 21, 2026new
Junyao Gao, Sibo Liu, Jiaxing Li et al.
Increasing the attention context length significantly improves streaming video generation quality, while incurring prohibitive memory and computational costs as the context grows. Existing methods attempt to capture this contextual information through attention selection or compression, but remain limited by bounded representations and lack explicit attention-level supervision to retain historical contributions to current queries. To address this, we introduce Attn-Forcing, a novel framework that preserves full-context attention contributions for streaming video generation under bounded memory and computation. We achieve this by introducing attention proxies that continuously approximate the attention responses induced by the historical context. Specifically, we implement the proxies as learnable visual latents and employ a Proxy Update Module to incorporate visual context from historical latents. Moreover, we propose an Attention Matching Loss to explicitly constrain the attention behavior of the learned proxies using full-context attention supervision. Extensive qualitative and quantitative experiments demonstrate that Attn-Forcing consistently improves streaming video generation quality and achieves state-of-the-art performance over existing baseline methods.
cs.CVOct 21, 2026new
Marek Wodzinski, Kamil Kwarciak, Daria Hemmerling et al.
A patient who has lost part of the skull vault needs an implant that fits the defect exactly, and such implants are still drawn by hand in CAD by specialists, which makes each case slow and expensive. The task can be considered as a shape completion problem, so pretrained completion models, and now the first shape completion foundation models, look like a way to remove that manual step. We audited three publicly released pretrained shape completion models, two of them foundation models, applied zero-shot to 239 cranial defects, some carved from public skull datasets and some real post-craniectomy cases. Each output was compared with simply returning the patient's own remaining bone, and tested for whether the predicted material could be placed without intersecting the bone that is still there. On the 210 carved defects none of the three improved on doing nothing, and on the 29 real cases only one did so beyond chance. Yet not one output, from any model in any configuration, was implantable: none managed both to stay clear of the bone the patient still has and to place more material than doing nothing, where a task-specific volumetric model, trained for this task and selected on these cases, was implantable on all but one. The output that most convincingly resembles a skull is no more placeable than the rest, and at these point densities implant thickness cannot be read off the emitted points at all, which is a property of the output representation rather than of any one model. On this evidence the shape completion foundation models we could obtain cannot be trusted and, used as released, are not usable in clinical practice. Substantial research towards trustworthy foundation models for shape completion is still needed.
cs.CL · cs.AIOct 21, 2026new
Yikai Zhao, Saurabh Pandey, Pradeep Kumar Misra
Enterprise agents act on documents. Before an agent can follow a policy, answer a compliance question or file a ticket, some ingestion layer must turn a Word file or PDF into tokens - and that layer is where an agent deployed in the wild silently loses information. We show that one layer is lost by every mainstream converter - the meaning an author encodes in how text looks - and that the loss makes agents confidently wrong rather than visibly incomplete. We build a task from 76 real policy documents: recover the requirements the author singled out, scored against the file's own OOXML properties by a deterministic matcher with no LLM judge. Formatting-blind ingestion collapses from micro F1 0.690 to 0.150 (docling) and 0.075 (plain text), a gap that replicates across six answering models from two vendors - five Claude models spanning the Opus, Sonnet and Haiku tiers, and OpenAI's GPT-5.6. An agent that cannot see the cue guesses, and docling's false-positive rate on requirements the author did not flag is four times ours. A span-level leave-one-out localises the loss to two cues - highlight (+0.302) and colour (+0.262), both Holm-adjusted p < 1e-4 - while bold, italic, monospace and strikethrough each sit within +/-0.003 of zero. The same asymmetry appears in PDF. Colour is also the cue no mainstream converter preserves and no parsing benchmark scores, and it is common: 32.8% of 6,460 documents carry a deliberate colour or highlight, 10.3% on an obligation. A pre-registered test shows that typing a cue for the agent adds +0.085 beyond merely preserving it (Holm-adjusted p = 0.049). We validate the labels against rendered page images through an independent modality (kappa = 0.72) and bound what wording alone can achieve (AUC 0.664). For agent builders the practical finding is that ingestion, not reasoning, is where this reliability is won or lost, and that the fix is cheap and measurable.
cs.CVOct 21, 2026new
Chenxu Zhou, Zhiyuan Yu, Yudong Jin et al.
This paper addresses the problem of creating immersive volumetric videos from sparse and low-resolution capture resources. Conventional dense multi-view reconstruction methods can produce high-quality volumetric videos, but their reliance on costly capture setups makes them impractical in real-world scenarios. Recent approaches use generative models to relax capture requirements, but still suffer from limited resolution and 4D inconsistency. To overcome these constraints, we propose RealMe, a unified framework built around a high-resolution video diffusion model that enhances limited captures into high-resolution (2K), 4D-consistent videos for volumetric video reconstruction. Specifically, RealMe constructs auxiliary data from capture videos and encodes them into latents along with layout-aware positional embeddings. The video diffusion model adapts factorized spatio-temporal attention to the multi-view capture layout, denoises these latents across viewpoints and time, and decodes them into enhanced videos to refine the 4D representation. This adaptation helps the model capture temporal dynamics and multi-view relationships, leading to more 4D-consistent generation. To reduce the data gap between generated videos and real captures, we apply a latent frequency loss that provides frequency-domain supervision in latent space. Extensive experiments on the public DNA-Rendering benchmark and our self-captured RealVV dataset demonstrate that RealMe is able to create immersive volumetric videos and outperforms state-of-the-art methods in visual fidelity and 4D consistency.
physics.chem-ph · cs.AIOct 21, 2026new
Yuhao Chen, Donald G. Truhlar, Xiao He
Empirical exchange-correlation functionals are fitted on fixed orbitals but used self-consistently, so better energies often come with worse densities. We show that the Brillouin residual of a trained functional on the Kohn-Sham orbitals of an accurate density controls, to first order, both the self-consistent density shift and the gap between fitted and self-consistent energies. Penalizing it requires no self-consistent-field solution during training. Trained on 6 energies and 11 CCSD densities, a neural meta-GGA surpasses SCAN in both atomization energies and densities of unseen molecules.
cs.LGOct 21, 2026new
Zhiyuan Ye, Xiangyu Zhou, Ji Qi et al.
Adversarial training derives updates from surrogate losses, but source-loss descent does not directly characterize an update's effect on adversarial margins. We introduce ExGrad (extrapolative gradient recomposition), which constructs clean-input, adversarial-input, and adversarial-weight-perturbed (AWP) gradients. Virtual steps measure their margin responses, and a globally solvable three-variable problem selects coefficients over a bounded affine domain without nonnegativity constraints. Across affine radii on CIFAR-10, negative clean coefficients rapidly predominate and increase in magnitude with the radius. Cross-zero interventions show that setting the clean coefficient to zero is insufficient: mild subtraction improves PGD-20, while stronger subtraction trades clean accuracy for further PGD-20 gains. In CIFAR-10 WRN-28-4 controls, none of the three candidate-removal variants matches full ExGrad. Projection diagnostics show that negative clean coefficients usually coexist with positive net clean-axis projections, although both decrease as the radius grows. Norm-matched controls indicate that update magnitude alone does not explain the moderate-radius gains. Experiments on CIFAR-10, CIFAR-100, and ImageNet-100 show radius-dependent improvements over AWP. On CIFAR-100, ExGrad at $R=0.8$ improves all five reported metrics, including clean and AutoAttack accuracy by 3.84 and 1.75 percentage points. On ImageNet-100, the same radius improves all four robust metrics over both AWP and PGD-AT. These results identify clean-gradient subtraction as a response-revealed update preference whose strength and clean-robust operating point are controlled by the affine radius.
cs.AIOct 21, 2026new
Meihong Jia, Yasemin Timar, Jonathan Taws
This study introduces GuideEvolve, a self-evolving agent framework incorporating expert guidance and three autonomous regulatory mechanisms, collectively termed the 3Ts: Think Fast and Slow, Throwback, and Termination by choice. The framework addresses critical limitations in extant self-evolving agent paradigms through dual innovations: (1) a task-specific expert-evaluation-driven evolutionary pipeline, and (2) the integration of 3Ts mechanisms to facilitate cognitive optimization, performance recovery via temporal rollback, and adaptive termination protocols. Empirical validation across the MGSM and MMLU benchmarks demonstrates that GuideEvolve attains performance metrics comparable to current state-of-the-art models (89.8% accuracy versus Gödel Agent's 90.6% on MGSM; 89.1% versus 87.9% on MMLU), while exhibiting notably reduced variance (25-59%) and enhanced robustness. Furthermore, the framework achieves a 100% evolution success rate, superior to the baseline's 82%, diminishes required optimization iterations by approximately 50% (from 30 iterations in baseline to 15.1 and 13.4 across tasks), and provides improved stability for production deployment.
cs.SE · cs.AIOct 21, 2026new
Sujoy Roychowdhury, Joseph H. R. Isaac, Jiangning Gao et al.
Understanding large, complex, and rapidly evolving code bases remains a core challenge for software engineering at scale, hindering onboarding, maintenance, refactoring, and impact analysis. Providing this information to a Large-Language Model (LLM) is another hurdle, as the LLM must understand not only the required code method but also its associated methods in the grand scheme of the repository. We propose Code Comprehensor, a hybrid retrieval-reasoning framework combining (i) a code dependency callgraph, (ii) CodeScribe, which uses code-aware LLMs to generate hierarchical documentation across repository call graphs, and (iii) CodeContrast, a CLIP-style dual-encoder model trained on CodeScribe output to learn aligned text-code embeddings. CodeContrast enables fast, semantically precise retrieval of relevant code-context neighborhoods, which serve as grounded evidence for an MCP-based LLM orchestrator answering domain-specific queries. Evaluation on OpenCV and a proprietary 5M-LOC repository shows that CodeScribe-derived supervision significantly improves cross-modal retrieval in text-to-code accuracy over baseline approaches. Richer textual supervision (hierarchical documentation, parent/sibling context, and LLM-generated synthetic questions) consistently enhances performance across both repositories. LLM-as-a-judge evaluation using the popular RAGAS of the Code Comprehensor output yields above 90% faithfulness scores.