Peer Revenue Past citations are no guarantee of future returns

Papers

Fresh from arXiv, typeset properly. Highlighted words trade on the market.

Computer Science

cs Watch Computer Science (sign in to use)

Tuesday 18 August

Computer Vision

GSToken: Geometry-Structured Gaussian Tokens for Compact 3D Medical Image Representation

Effective segmentation of multi-modal MRI is central to improving neural network accuracy in brain tumor recognition. Existing methods typically compress 3D volumes into token sequences via fixed patch encoding or learned attention pooling (e.g., TokenLearner). However, these compression schemes discard explicit spatial shape information; the resulting tokens convey no notion of lesion morphology or spatial extent. Meanwhile, end-to-end evaluation entangles a tokenizer's information retention with the reconstruction capacity of the downstream decoder, and the lack of a unified capacity contract across methods makes performance differences difficult to attribute. In this paper, we introduce Gaussian tokens to multi-modal brain tumor segmentation for the first time: each token carries not only a semantic feature but also a learned 3D center, anisotropic scale, and orientation, endowing the representation with explicit geometric support at negligible parameter cost. We further propose a frozen-token utility evaluation protocol: the trained tokenizer is frozen, its output is cast into a fixed-capacity serialized contract, and a shared lightweight Transformer probe independently measures each tokenizer's retained information under strictly matched conditions. Multi-seed paired statistical testing shows that GSToken consistently and substantially outperforms capacity-matched adaptive baselines under frozen probing, with uniform advantages across all tumor sub-regions, surface, and distance metrics. These results demonstrate that explicitly encoding spatial geometry within tokens significantly improves the information density of volumetric representations, offering a new design principle for compact 3D medical image representation and downstream reading.
Read the full abstract

Robotics

Prism-GRPO: Faster VLA Policy Optimization via Splitting Same-outcome Groups

GRPO is increasingly used for reinforcement learning of vision-language-action (VLA) policies because, unlike PPO, it does not require training a critic. This simplification comes with a sampling cost: group-relative advantages require multiple rollouts from each scene. Under binary success rewards, groups whose rollouts all succeed or all fail have zero advantage and are discarded by dynamic sampling. These groups are especially common early in training, when most rollouts fail, wasting much of the expensive robotic rollout budget. We introduce Prism-GRPO, which augments binary outcome reward with a weighted trajectory-level execution-quality score. By splitting same-outcome groups into a quality spectrum, Prism-GRPO recovers training signal while ensuring that every success still outranks every failure. Quality scores can be derived from simulator contacts, executed actions, or visual observations, avoiding task-specific progress rewards. We prove that Prism-GRPO never increases the probability that a sampled group is discarded for having zero advantages, and derive a gradient-alignment condition under which its combined update remains a local ascent direction for task success. Across four RoboTwin tasks spanning different horizons and coordination patterns, Prism-GRPO improves success and quality at matched rollout budgets and reaches target success rates with up to 56% fewer rollouts. It also suppresses a reward-hacking shortcut, with the cleaner behavior transferring under direct deployment to a real robot. Through ablations, we show consistent gains across contact-, smoothness-, and VLM-derived quality signals.
Read the full abstract

All authors Zeyun Deng, Yuzhe Lu, Yawei Wang, Linbo Liu, Qing Ping, Han Ding, Guande Wu, Panpan Xu, Jun Huan

Computer Vision

TF-CADE: Foreground-Concentrated Text-Video Alignment for Zero-Shot Temporal Action Detection

Zero-Shot Temporal Action Detection (ZSTAD) aims to lo- calize and recognize action instances from unseen action categories in untrimmed videos. Although existing meth- ods have shown effectiveness by advancing architectural text-video alignment, they still struggle with capturing se- mantic distinctions between action classes, resulting in text- irrelevant predictions. To address this issue, we propose a Text-Foreground Concentrated Alignment for zero-shot temporal action DEtector (TF-CADE) that explicitly aligns textual information with action-relevant foreground regions. Specifically, we introduce Action Concentrate Aggregation (ACA), which extracts action concentrate scores to aggregate temporally informative video segments into a foreground- weighted video embedding. This foreground concentrated alignment enhances the semantic consistency between text and video features and improves inter-class discriminabil- ity. In addition, a Certainty-based Confidence Re-weighting (CCR) strategy refines per-snippet confidence scores by lever- aging foreground-aware similarity, effectively suppressing irrelevant action classes during inference. Extensive evalua- tions show that our TF-CADE not only achieves state-of-the- art performance under in-distribution settings but also excels in cross-dataset generalization to unseen action classes.
Read the full abstract

Computer Vision

TEAMS: Text-prompted spatiotEmporal dual-heAd Mamba Snake

Deep snake is a promising family of instance segmentation methods that accurately predicts object-level contours, thereby overcoming common pixel-level misclassification issues such as mask cavities and jagged edges in semantic segmentation approaches. However, existing deep snake methods face challenges in handling complex morphological variations, accurately capturing fine-grained organ details, and correcting base detection errors. To mitigate these limitations, we propose a cohesive Text-prompted spatiotEmporal dual-heAd Mamba Snake (TEAMS), a novel vision-language Mamba snake framework with three key innovations: (1) A Spatiotemporal Snake Evolution Strategy (SSES) is introduced to tackle complex morphological variations by capturing bidirectional spatial dependencies along the snake contour and temporal dynamics across evolution steps in a state space model. (2) A Contour Morphology-Aware Mamba (CMAM) is proposed to quantify local contour morphologies to modulate the structured attention mask in the Mamba2 SSD dual form, which extends Mamba's capability to perceive the relative importance of its input sequence elements for better delineation of fine-grained organ details. (3) A Text-prompted Collaborative Dual-Head Snake (TCDHS) is designed to incorporate cues from textual prompts and transfer the evolved contour information to the base detection head, which enhances the deep snake workflow and mitigates wrong detections. Comprehensive evaluations on five datasets covering different organs and imaging modalities demonstrate that TEAMS outperforms existing semantic and deep snake segmentation methods (e.g., relative mDice/mBF improvements of 6.9%/9.1% in a spinal dataset), underscoring its potential as a reliable tool across diverse medical image segmentation scenarios.
Read the full abstract

All authors Ruicheng Zhang, Jianhui Lei, Kaiwen Shen, Haowei Guo, Jun Zhou, Bin Chen, Mengtang Li, Shen Zhao, Shuo Li

Computer Vision

SPVC: Structured and Panoptic Video Fixing for Cross-Dataset Driving Scene Rendering

Driving scene reconstruction and rendering, especially with 3D Gaussian Splatting, has become an important component of autonomous driving simulation. However, rendered views often degrade under extrapolated ego trajectories and scene edits, producing blurry structures, temporal flicker, and foreground-background misalignment. Existing refinement methods are commonly designed for a specific setting, such as image-level novel-view repair or object-editing correction. In this paper, we introduce SPVC, a structured and panoptic video fixing framework for cross-dataset driving scene rendering. The name summarizes four design principles. (1) Structured fixing denotes the use of explicit spatial conditions, including camera pose, 3D bounding boxes, and HD maps, to guide the repair process and reduce uncontrolled hallucination. (2) Panoptic fixing refers to correcting both background rendering artifacts, such as distorted roads, buildings, and lanes, and foreground vehicle artifacts introduced by scene editing, such as inconsistent object appearance. (3) Video fixing means that the model operates on driving sequences rather than isolated frames, allowing temporal cues to be used during artifact correction. (4) Cross-dataset fixing means that a single shared network is trained and applied across multiple driving datasets, reducing the need for dataset-specific or scene-specific fixers. Concretely, we construct paired degraded-clean training data by simulating under-constrained 3DGS rendering and foreground vehicle insertion artifacts, and train a two-stage controllable video diffusion model that first addresses video-level appearance and then refines scene layout with structured controls.
Read the full abstract

All authors Gen Li, Shu Han, Yun Xi Qiao, Hua Chen, Xuyang Dai, Bohan Li, Hao Zhao, Chaojian Li

Game Theory

Self-Bounding Regret Matching+ in Potential Games and Product-Simplex Optimization

Regret matching+ (RM+) is parameter free, scale invariant, and central to large game solving, but its only general individual-regret guarantee grows as $\sqrt{T}$. A recent ICLR result used this envelope to prove that RM+ reaches an $\epsilon$-stationary point of a smooth objective over a product of simplices in $O(\epsilon^{-4})$ iterations, or $O(\epsilon^{-8})$ from the standard zero initialization. We give an exact one-step conservation law for RM+. It states that forward utility gain pays for both squared state motion and growth of the regret-state norm. Norm growth is at most $\sqrt{m-1}$ times forward gain for $m$ actions, and the coefficient is sharp. This yields four results for unmodified RM+. Its regret on any utility path is controlled by centered temporal variation. Its regret is uniformly bounded under alternating play in every finite exact potential game, resolving an open question and making squared activation gaps summable. Both certified lazy and ordinary cyclic play attain an $\epsilon^{-2}$ exponent. On any smooth, possibly nonconcave simplex objective, RM+ finds an $\epsilon$-KKT point in $O(\epsilon^{-2})$ iterations. Most broadly, for a smooth objective over an arbitrary product of simplices, cyclic block RM+ attains the same $O(\epsilon^{-2})$ exponent from arbitrary initialization, with an explicit trajectory-dependent constant. The proof controls the finite objective loss caused by low-state blocks and then self-bounds every block state and the total squared path length. Complete proofs cover zero states, sharpness, common-profile stationarity, and robust gain dominance. Oracle-normalized diagnostics compare RM+ with predictive and smooth extra-gradient variants on graphical potential games and dense nonconvex objectives.
Read the full abstract

Robotics

Bi-Layer Ant Colony Optimization for Multi-Robot Task Allocation and Routing in Delivery Applications

This paper addresses the multi-robot task allocation (MRTA) problem, which is essential for delivery and logistics applications. Our approach first defines a new cost function that transforms the MRTA into a unified optimization problem capturing both task assignment and routing. A bi-layer ant colony optimization (ACO) algorithm is then introduced, integrating two interdependent decision layers within a single colony process to solve the problem. This hierarchical framework enables simultaneous optimization of task allocation and route planning across multiple robots. Comparative experiments with mixed-integer linear programming (MILP) and particle swarm optimization (PSO) demonstrate that the proposed bi-layer ACO achieves the shortest total travel distance and fastest completion time across all task sizes. Specifically, it reduces total travel distance by up to 17.7% and completion time by nearly 20% compared with baseline methods. These results confirm the efficiency, scalability, and reliability of the proposed bi-layer ACO for multi-robot delivery tasks.
Read the full abstract

All authors Le Na Nguyen, Thanh Long Nguyen, Thanh Thao Ton Nu, Quan Le, Manh Duong Phung

Computer Vision

Spectral Gradient Orthogonalization Improves Differentially Private Training at Scale

Differentially private training adds isotropic Gaussian noise to clipped gradients, corrupting every singular direction equally. In vision models, where spatial correlation concentrates gradient energy into a low-rank subspace, most of this noise falls in directions that carry little signal. Spectral gradient orthogonalization via polar decomposition is introduced as a post-processing step that recovers directional signal from the noisy gradient's low-rank structure at zero additional privacy cost. A phase transition governs the utility of this approach: orthogonalization improves accuracy only when the per-direction spectral signal-to-noise ratio (SNR) suffices for singular vector recovery; in low-SNR regimes, the directional bias of the gradient is replaced by a nearly random orthogonal update, and the transformation is harmful. The recovery threshold is determined by the spectral gap of the gradient and is surpassed at large batch sizes. Empirically, the benefit scales with model capacity: spectral orthogonalization achieves a +20.9% improvement over DP-SGD on WRN-28-10 (B = 4096) and +14.9% on ResNet-18, while reducing inter-run variance by a factor of two to three. In the fine-tuning regime, spectral orthogonalization matches the stability of DP-Adam while maintaining a first-order memory footprint. Combining spectral with temporal denoising yields 50.3% on CIFAR-10 (epsilon = 4), the highest accuracy in any tested configuration. These gains are specific to moderate-to-high-SNR regimes such as large-batch training of higher-capacity models. Small-batch or low-SNR settings are better served by DP-SGD or temporal denoising.
Read the full abstract

Computer Vision

REChart: Reasoning-Efficient Chart Editing with Large Reasoning Models

Chart editing requires inferring and modifying visualization code from a reference chart image based on an editing instruction, challenging fine-grained visual reasoning, instruction following, and executable code synthesis capabilities of MLLMs. Large reasoning models (LRMs) with extended Chain-of-Thought (CoT) reasoning are suitable for tackling such complex multimodal tasks. However, our preliminary study reveals an “inverted-U” relationship between reasoning length and chart-editing performance: Excessive reasoning often leads to “overthinking,” where models drift toward hallucinated visual details or get stuck in redundant reasoning loops. To address the gap, we introduce REChart, a two-stage training framework that provides process-level supervision over intermediate reasoning steps, improving both editing fidelity and reasoning efficiency. First, we synthesize 200k high-quality reasoning trajectories for supervised fine-tuning from a large image-instruction-code pool, using a role-specialized agentic Reason-Score-Refine workflow that iteratively refine the chart code toward higher quality. Second, we optimize the model via reinforcement learning with two complementary rewards: a fidelity reward evaluating code correctness, visual fidelity, and structural consistency, and an efficiency reward that assigns each rollout a random thinking budget, truncates the reasoning process, and credits the final reasoning segment according to its contribution to the output. On the ChartEdit and ChartMIMIC benchmarks, our model achieves state-of-the-art chart-editing performance among open-source models of comparable scale, while mitigating overthinking and reducing average reasoning token usage by 79.0% under a maximum thinking budget of 16,384 tokens compared with the base model.
Read the full abstract

All authors Yuanbang Liu, Chenxi Ruan, Yihan Hou, Qiong Luo, Wei Zeng

Machine Learning

GUPO: Gradient Uncertainty-aware Policy Optimization for Post-Training Large Language Models

Group Relative Policy Optimization (GRPO) has become a widely used approach for post-training Large Language Models (LLMs) for reasoning. In GRPO, the group gradients induced by different queries within the same mini-batch are directly averaged to form the policy update. However, these group gradients can point in conflicting directions. Our empirical analysis suggests that group-gradient conflicts tend to be associated with less effective policy updates, motivating the need for a reliable aggregated update direction under such conflicts. Standard GRPO aggregation treats the realized group gradients as deterministic contributions and does not account for differences in their reliability during aggregation. To address this issue, we propose Gradient Uncertainty-Aware Policy Optimization (GUPO), which models each group gradient as a random variable under a Bayesian formulation and estimates its probability distribution. GUPO then derives gradient uncertainty using a Dirichlet-based formulation and uses it to calibrate the contribution of each group gradient during aggregation. Extensive experiments on multiple benchmarks demonstrate the effectiveness of GUPO.
Read the full abstract

All authors Peizheng Guo, Jianqi Zhang, Xingyu Zhang, Yun Fan, Jiahuan Zhou, Changwen Zheng, Wenwen Qiang

Human-Computer Interaction

The Oracle of Chemnitz: An interactive art installation to reanimate old things in a garage featuring a rotary phone

Garages have a long tradition of tinkering, creativity and innovative change. School of Garage, a participatory artistic summer school project in Chemnitz, the European Capital of Culture 2025, took up this tradition and turned old Eastern Bloc garages into temporary ateliers for collaborative making and discussion. In our HackLab garage we conceptualized and created the Oracle of Chemnitz within one week. It gives a place filled with history back its stories. It is an interactive installation of artifacts from the past typically found in garages: an old typewriter, radio, desk, tires, mixer and a rotary-dial telephone. Each got a name, personality and story to tell. The phone rings when a visitor approaches. Once answered, it asks for name and month of birth before a story about a device is told, along with hints to other places in the city. Around 2,700 visitors interacted with the system over three months.
Read the full abstract

Computer Vision

MoE-ViE: Mixture of Experts Vision Encoder for Efficient Image and Video Understanding

Vision encoders are a critical component of vision-language models, and scaling their capacity effectively improves performance. However, dense scaling increases compute cost and inference latency. Mixture-of-Experts (MoE) architectures offer a compelling alternative, having enabled efficient scaling in LLMs, yet the MoE design space for CLIP-style vision encoders remains underexplored at State-of-the-Art (SOTA) levels. In this work, we systematically study MoE designs for vision encoder scaling and find that fine-grained MoE topologies yield substantial gains over both dense and standard MoE counterparts. We further propose an auxiliary-loss-free balancing variant for better expert utilization, and design a specialized MoE kernel to mitigate inference latency overhead. To enhance video capabilities while preserving image knowledge, we introduce frame-level distillation paired with a novel freezing mechanism. We pretrain a series of Mixture-of-Experts Vision Encoders (MoE-ViE) across a range of sizes, all consistently outperforming their dense counterparts. Our largest model matches the zero-shot performance of a SOTA encoder 1.7x its size at 76% of its latency. When aligned with an LLM, MoE-ViE surpasses all compared encoders on image and video benchmarks, including those with up to 5x more activated parameters. Code is available at https://github.com/facebookresearch/moe_vie.
Read the full abstract

All authors Bonan Zhang, Shiyu Dong, Quan Hung Tran, Katharina Gschwind, Shuqi Yang, Sijia Chen, Adel Ahmadyan, Seungwhan Moon, Lu Zhang, Ahmed Kirmani, Babak Damavandi, Anuj Kumar

Software Engineering

COMMITGUARD: Differential Slice Fuzzing for Commit-Induced Bug Detection

Modern software systems evolve through frequent commits that implement bug fixes, features, and security patches. Although code review and testing are widely used to check these changes, they often provide limited assurance for memory-safety issues. Code reviewers may miss subtle boundary, lifetime, or initialization errors, while existing tests may not exercise the specific paths affected by a commit. Fuzzing is effective at exposing such bugs, but applying it to every commit remains impractical because whole-program fuzzing is expensive, requires suitable harnesses, and may still fail to reach the code changed by a commit. In this paper, we introduce COMMITGUARD, a commit-aware differential slice-based fuzzing approach for verifying code changes. The key insight behind COMMITGUARD is that the pre-commit version of a modified function can serve as a behavioral baseline for interpreting bugs found after the commit. For each target commit, COMMITGUARD identifies modified functions, extracts compilable code slices from both the pre-commit and post-commit versions, and fuzzes the paired slices independently. It then compares sanitizer reports across the two versions and reports bugs that emerge only in the post-commit version as candidate commit-induced bugs. We evaluate COMMITGUARD on 300 commits from openSSL, libpcap and leptonica. Slice fuzzing initially produces 518 sanitizer reports across these commits. By comparing pre-commit and post-commit slices, COMMITGUARD narrows this large output to 7 candidate commit-induced bug reports that require manual triage. Manual validation confirms 5 of these reports as real bugs that were fixed by developers of the examined projects after we reported them, while only 2 reports were classified as false positives. COMMITGUARD analyzes a commit in 32.4 minutes on average and achieves 75.36% average coverage of modified functions.
Read the full abstract

All authors Aniruddhan Murali, Noble Saji Mathews, Mahmoud Alfadel, Meiyappan Nagappan

Computation and Language

An Investigation of Translationese in the Generations of Multilingual Large Language Models

Text which has been translated from another language tends to carry with it evidence of translation$\unicode{x2014}$hence, it is often referred to as $\textit{translationese}$. Multilingual large language models (MLLMs) generate text in a variety of languages. However, it is still unclear if MLLMs' generations resemble internal translation (from English or, potentially, other languages) and, thus, result in translationese. Here, we ask the following research questions: (1) Does text generated by MLLMs resemble translationese? (2) How does translationese produced by MLLMs differ from translationese produced through direct translation? We leverage established indicators of translated text to evaluate text generated by state-of-the-art MLLMs in five languages, comparing to both non-translated and human-written baselines in order to isolate translationese from other kinds of interference. Through the use of high-accuracy classification models, analyses of variance on individual linguistic features, and the collection of human annotations in a subset of two languages (German and Spanish), we assess the translationese content of MLLM generations and examine the key features that distinguish MLLM-generated text from typical translation-related interference.
Read the full abstract

All authors Maria Valentini, Téa Wright, Julisa Granados, Eliana Colunga, Katharina von der Wense

Computer Vision

To Remove or Not to Remove Clouds: A Comparative Analysis and Fusion of Raw SAR and Synthetic NDWI for Overcast Water Segmentation

Persistent clouds blind optical satellites during floods. While Synthetic Aperture Radar (SAR) penetrates clouds, its raw data is noisy and lacks clear contrast. To mitigate this, recent studies utilize deep learning models to translate SAR into cloud-free synthetic optical imagery for downstream tasks like water body segmentation. However, because raw SAR is the original source for both of these operations, a critical methodological dilemma arises: during complete overcast should segmentation models process the raw SAR directly, or rely on a translated synthetic Normalized Difference Water Index (NDWI) proxy? This study resolves the debate by demonstrating that synthetic NDWI yields better results, as the translation process acts as a powerful filter against radar noise. This raises a natural second question: what if we utilize both? Building on our findings, we introduce a Combined Framework that integrates both raw SAR and synthetic NDWI into a unified model. By fusing the sharp physical boundaries of raw SAR with the high contrast of synthetic NDWI, this hybrid approach consistently outperforms all standalone methods.
Read the full abstract

Software Engineering

SNIPTEST: Fuzzing Multi-Level Code Slices for Validating Vulnerabilities

Modern software systems are increasingly complex, and static analysis tools are commonly used to identify potentially vulnerable code by issuing warnings. However, these warnings often require manual inspection to confirm whether the reported issues are real, making the process time-consuming and error-prone. Directed fuzzing has emerged as a powerful automated technique to validate the warnings. However, applying it to the entire project in response to each warning is computationally infeasible, often requiring days of execution to achieve only incremental improvements in code coverage. We present SNIPTEST, an execution-based warning triage framework that generates and fuzzes compiled code slices centered around static-analysis warnings. Rather than proving exploitability in the full program, SNIPTEST provides evidence about how a warning behaves under progressively expanded sliced execution contexts. It employs a layer-by-layer slicing strategy, incrementally expanding context around the target location to validate potential vulnerabilities with increasing precision. We evaluate SNIPTEST on a benchmark of 97 true vulnerabilities and 97 false alarms across three real-world projects. SNIPTEST produces Possible True Positive evidence for 53 of 97 confirmed vulnerabilities (54.6%) by triggering the corresponding bug oracle consistently across all three analyzed slice levels, while the remaining cases are unreachable. Particularly, in 40.2% of these cases, it exploits the vulnerability along the observed execution path, matching the top three stack frames. On the 97 confirmed false alarms, SNIPTEST produces Possible False Positive evidence for 54 cases (55.6%) by reaching the warning without triggering the bug oracle, but misclassifies 28 cases (28.8%),and the remaining cases are unreached. Finally, we demonstrate the practical relevance of SNIPTEST by identifying CVE-2025-11964.
Read the full abstract

All authors Aniruddhan Murali, Nobble Saji Mathews, Mahmoud Alfadel, Meng Xu, Meiyappan Nagappan

Computer Vision

Noisy group neurons with synchronous resetting for high-performance spiking neural networks

Spiking neural networks (SNNs), characterized by bio-inspired neuronal dynamics and event-driven communication, have attained significant progress in recent years. Nevertheless, training deep SNNs remains challenging due to spatiotemporal information loss and gradient mismatching. To simultaneously address these issues, we propose a noisy group neuron (NGN) model, which incorporates population-level synchronous resetting and neural stochasticity as fundamental computational mechanisms. We then develop the NGN method as a framework that combines the NGN model with backpropagation learning based on mean-field dynamics. We demonstrate the advantages of the NGN method through theoretical analysis and experimental validation on CIFAR-10, CIFAR-100, Tiny-ImageNet, DVS-Gesture, N-Caltech101, and CIFAR10-DVS. The proposed approach achieves an accuracy of 87.35% on CIFAR10-DVS within 10 inference time steps. These results support NGN as a practical approach to high-performance neuromorphic computing.
Read the full abstract

All authors Yajie Zhai, Yanmei Kang, Meng Li, Zigang Huang

Artificial Intelligence

LEGO-RL: Harness-Native Reinforcement Learning for Coding Agents

Reinforcement learning for coding agents increasingly relies on long-running agent harnesses to manage tool integration, repository contexts, and execution feedback. However, the native execution environments of these harnesses are inherently misaligned with policy-gradient training: environmental crashes and reward hacking corrupt outcome signals, while train-inference discrepancies decouple rollout behavior from policy updates. To address this, we present LEGO-RL, a framework that bridges native coding-agent harnesses with scalable policy-gradient optimization without modifying their internal control flow. LEGO-RL is built upon three pillars: (1) faithful optimization via in-process LLM proxying that captures raw generation streams for token-level alignment and robust trainer-side log-probability recomputation, even under harness-side compaction or re-serialization; (2) reliable execution via scalable sandbox orchestration featuring image caching and stage-wise defenses to mitigate reward hacking; and (3) observable training through an integrated plugin that automates validation and monitoring, paired with a Live UI for granular trajectory diagnostics. We evaluate LEGO-RL by training the sparse MoE model Qwen3.5-35B-A3B with GSPO across three native coding-agent harnesses. LEGO-RL improves Qwen3.5-35B-A3B across OpenHands SDK (64.0% to 70.4%), Claude Code (62.4% to 68.2%), and OpenCode (57.2% to 66.6%) on SWE-bench Verified, while maintaining a rollout-training probability correlation above 0.99.
Read the full abstract

All authors Yiming Du, Yuxin Jiang, Tao Yuan, Jianbo Dai, Shaowei Wang, Jierun Chen, Chaofan Tao, Xianzhi Yu, Lifeng Shang, Kam-Fai Wong, Xiaohui Li, Haoli Bai

Graphics

Six Ways to Draw Vangers with WebGPU: Real-Time Rendering of Editable Multi-Layer Height Fields

Terrain level-of-detail is measured almost exclusively on digital elevation models: single-valued, smooth at the sampling scale, sampled from real topography. Game terrain is often none of these. We compare six rendering methods - height-field ray marching, voxel-accelerated ray marching, sliced proxy geometry, per-sample bar rasterization, compute scattering, and a fitted triangle mesh - implemented in a single engine over a single data path, on the hand-authored multi-layer terrain of Vangers (1998), scored against a CPU ray cast of the same source data. Every method must preserve the two solid intervals available at a ground sample, render at interactive rates, and reflect local terrain destruction without reloading the level. These constraints rule out treating caves as decoration or amortizing a static preprocessing step over an immutable map. From the original game's top-down camera the six methods look interchangeable. At eye-level horizons they do not: point scattering loses coverage, slicing bands, and an over-simplified mesh can miss a wall. At the selected quality settings a greedy triangulated irregular network (TIN) has the lowest mean frame time on every device we measured, but the fit cost is set by the second layer rather than by floor relief, and making that mesh editable retains 319 MiB of GPU geometry and 535 MiB of CPU triangulation. All six implementations use the same native wgpu / WebGPU API and canonical WGSL. We release the engine, the harness, and a one-command measurement protocol.
Read the full abstract

Computer Vision

GeoWeaver: Accurate Long-Sequence 3D Reconstruction via Hierarchical Geometric Assembly

Long-sequence 3D reconstruction from RGB videos requires both accurate local geometry and globally consistent camera motion. Feed-forward models provide strong depth and pose predictions, but their memory cost prevents joint inference over long sequences. Chunk-wise processing improves scalability, yet independently predicted chunks often exhibit scale drift, pose errors, and point-cloud misalignment. We present GeoWeaver, a unified framework comprising a Geometric Prior Model (GPM) and Test-Time Adaptation (TTA). The GPM predicts chunk-wise depth, confidence, and camera parameters as adjustable geometric priors. TTA then performs sequential initialization, global chunk-level Sim(3) alignment, and coarse-to-fine refinement of camera poses, affine depth corrections, and intrinsics. Dense correspondences provide adjacent, cross-chunk, and long-range constraints, while a robust CDF-style objective jointly optimizes weighted 2D reprojection and 3D consistency residuals. This design preserves local geometric accuracy while correcting accumulated pose, scale, depth, and calibration errors. Experiments across diverse long-sequence benchmarks demonstrate improved camera accuracy, global consistency, and point-cloud quality. Ablations verify the contribution of each adaptation stage, and applying the same TTA procedure to different geometric prior models consistently improves their trajectory estimates, demonstrating that GeoWeaver is not tied to a specific GPM.
Read the full abstract

All authors Tinghao Jiang, Sheng Tang, Shengzhe Wei, Juntong Fang, Weiqi Zhang, Junsheng Zhou, Zesong Li

Distributed Computing

Generalizing and accelerating consistency checking for non-transactional distributed storage systems

Linearizability checkers check if an operation history, observed by concurrent clients, is linearizable. They are used in testing distributed storage systems, and use the classic Wing-Gong (WG) linearizability checking algorithm. In this paper, we generalize the WG algorithm to make linearizability checkers more versatile: we can check other non-transactional consistency guarantees, like ordered sequential consistency provided by Zookeeper. Equipped with this generalization, we can also check for system-specific consistency guarantees that introduce additional ordering constraints over operations in a history, as per the system's specification. Our experiments with 8 distributed storage systems show that checking for system-specific consistency guarantees is easy to realize, reduces false negatives in testing, helps debug consistency violations, can be up to 370x faster, and can scale to more concurrent clients within the same checking time budget. We report 6 new consistency violation bugs, out of which 5 could not be found with existing consistency checkers.
Read the full abstract

All authors Kotikala Raghav, Aman Hassan, Brian Sajeev Kattikat, Patel Jay, RSRS Santhosh, Abhilash Jindal

Robotics

MANIGUARD: A Benchmark and Data Suite for Specification-Grounded Safety Evaluation and Improvement of Robotic Manipulation

Foundation-model policies for robotic manipulation are advancing rapidly on task success, but rigorous evaluation of whether they succeed safely is still lacking. We introduce ManiGuard, a specification-grounded framework for evaluating and improving the safety of foundation-model manipulation, comprising the ManiGuard-Bench task suite and a paired safety-annotated trajectory-generation pipeline. ManiGuard-Bench organizes six contact-rich household task families into 200 locked base tasks along a skill $\times$ constraint taxonomy, with safety specified independently of task success. Each task is evaluated under one in-distribution and four single-axis out-of-distribution perturbations that hold the safety specification fixed, giving 1,000 locked scenarios. Every rollout is runtime-checked by LTL$_f$-grounded automaton monitors over physics-grounded predicates rather than learned classifiers or LLM judges, in simulation and on a physical Franka platform. The pipeline pairs an automated motion-planning generator with human teleoperation, annotated by the same per-step monitor, and directly supports safety-aware fine-tuning; we release 8,000 safety-annotated demonstrations, 40 per base task. Benchmarking zero-shot and fine-tuned VLAs across more than 23,000 rollouts, we find: (i) safety must be evaluated independently of task success, as 6-21% of successful rollouts violate the specification; (ii) fine-tuning on our suite raises safe task completion from near zero to 7.5-29.8% and engaged-and-safe behavior from 16-40% to 51-72%; but (iii) a gap remains that scaling demonstrations does not close, with 21-42% of engaged rollouts still violating, two of six families below 2% safe success for every policy, and these failures persisting under distribution shift and on hardware.
Read the full abstract

All authors Yiyan Peng, Philip Wang, Simon Sinong Zhan, Yiqi Lyu, Zhenyang Ni, Jixin Yan, Fiorelli Wong, Ruochen Jiao, Hang Yin, Xinyu Cao, Huajie Shao, Manling Li, Ruohan Zhang, Qi Zhu

Data Structures and Algorithms

Maximum Flow Without the Outer IPM

We show that the balancing weights technique of Li (2026) actually produces an approximate *pseudo-circulation* of a directed, capacitated graph in $m^{1+o(1)}$ time. Together with standard flow techniques, we obtain an $m^{1+o(1)}$ time maximum flow algorithm that avoids the interior-point method framework of recent almost-linear time algorithms (Chen et al. FOCS 2022, van den Brand et al. FOCS 2024).
Read the full abstract

Information Theory

Universal CKM for Environment-Aware Wireless Networks: Enabling Cross-Device and Cross-Task Channel Knowledge Transfer

Channel knowledge map (CKM) is a promising technology for environment-aware sixth-generation (6G) wireless networks. However, existing CKMs are tightly coupled with wireless devices and downstream tasks, which limit their scalability and reusability in wireless networks. To address these limitations, this article proposes the concept of universal CKM (uCKM) as a foundational wireless environment prior, which aims to enable cross-device and cross-task channel knowledge transfer for environment-aware wireless networks. We first revisit the representative CKMs and discuss their limitations. Then, the uCKM-enabled new paradigm for environment-aware wireless networks is introduced, and its benefits are highlighted from the perspectives of uCKM construction and utilization phases, for which we propose the visions of “All for uCKM” and “uCKM for All”, i.e., the data acquired by all devices and tasks should contribute to the construction of uCKM, while the constructed uCKM, in turn, can be utilized to support all devices and tasks. Subsequently, we discuss the main challenges of uCKM and propose potential solutions. Last, we provide simulation results to demonstrate the feasibility and performance gains brought by uCKM and outline future research directions.
Read the full abstract

All authors Haiquan Lu, Yong Zeng, Cheng-Xiang Wang, Xiqi Gao, Rui Zhang

Artificial Intelligence · Cross-listed from q-bio.QM

Leveraging generative hallucination and biophysics-informed modeling for unified biomolecular sequence-structure co-design

Biomolecular design underpins applications from molecular recognition to therapeutics and synthetic biology, yet de novo interaction design remains challenging-especially for DNA/RNA, underexplored non-protein modalities with scarce, heterogeneous complex data and sharper geometric and chemical constraints. We introduce MCTH (Monte Carlo Tree Hallucination), an inference-only framework that casts all-atom sequence-structure co-design as uncertainty-aware planning over hallucinated states from pretrained folding and inverse-folding models, with optional biophysical control within the same decision loop. MCTH treats these models as frozen black-box operators and uses Monte Carlo Tree Search to allocate a fixed inference budget across competing design trajectories, incorporating model confidence and uncertainty, as well as cross-expert consensus/disagreement when multiple predictors are available. Across protein-RNA, protein-DNA, protein-protein, and protein-ligand design, matched-budget experiments show that adaptive search improves over simpler sampling and cycling strategies, while held-out AlphaFold3 and Chai-1 evaluations demonstrate transfer beyond the search-time oracle. MCTH provides a shared planning layer across modalities while allowing task-specific folding, inverse-folding, and biophysical modules, requiring no fine-tuning or backpropagation through component models.
Read the full abstract

All authors Xuefeng Liu, Mingxuan Cao, Xiao Luo, Songhao Jiang, Tobin Sosnick, Jinbo Xu, Louis Maher, Rick Stevens

Computation and Language

PTXBench: Benchmarking and Adapting LLMs for GPU Kernel Optimization with Architecture-specific PTX

We introduce PTXBench, a benchmark for evaluating and adapting large language models (LLMs) to use architecture-specific PTX for GPU kernel optimization. PTXBench measures functional correctness, whether selected target instructions execute at runtime, and speedup over frontier libraries across GEMM and attention workloads on H100 and B200 GPUs. Our evaluation shows that architecture-specific PTX capability remains uneven: success rates fall substantially on complex attention backward workloads, and executing the target instructions does not necessarily translate into competitive performance. No evaluated model consistently matches frontier libraries across the suite. We further adapt Qwen3.6-27B using supervised fine-tuning. Repair-conditioned training improves several tasks, but generalization remains uneven; data coverage, balance, and the quality of the reasoning teacher matter in addition to dataset size. PTXBench provides an auditable testbed for measuring and improving LLMs' ability to exploit evolving GPU architectures.
Read the full abstract

All authors Genghan Zhang, Yixin Dong, Chengze Fan, Zhichen Zeng, Yueming Yuan, Shaowei Zhu, Kunle Olukotun

Cryptography and Security · Cross-listed from math.PR

On the Pseudo-Mixing of Kac's Walk

Motivated by a conjecture of Vaikuntanathan and Zamir, we study the pseudo-mixing of Kac's walk on $\mathrm{SO}(n)$: whether short trajectories are indistinguishable from Haar measure by low-complexity tests. We prove that the first $k$ columns mix in Wasserstein distance in $O(n(k+\log n)\log n)$ steps for fixed accuracy, resolving a conjecture of Oliveira. Combining this with a representation-theoretic variance bound, we show that if $T=\omega(nk(k+\log n)\log n)$, then every degree-$k$ polynomial normalized to have unit Haar variance has expectation under the $T$-step law within $o(1)$ of its Haar expectation. As an application, we show that this pseudo-mixing estimate can be used to prove the effectiveness of a fast Johnson–Lindenstrauss transform with the usual target dimension.
Read the full abstract

Machine Learning

Integrating Novelty and Surprise for Experience Prioritization and Exploration in Image-Based Reinforcement Learning

Sample efficiency is a central challenge in reinforcement learning (RL), particularly in image-based domains where agents must learn from high-dimensional visual inputs. Traditional sampling often relies on random or suboptimal experience selection, leading to redundant updates and slow learning. Improving efficiency requires mechanisms that prioritize informative experiences while also encouraging effective exploration. Prioritized Experience Replay (PER) addresses part of this challenge by reusing high-value transitions, while intrinsic rewards promote the exploration of novel or uncertain states. However, their integration has not been extensively studied. This paper introduces Novelty and Surprise Prioritized Experience Replay (NSPER), which uses novelty to capture underrepresented states and surprise to expose gaps in the agent's understanding of the environment. We further extend this with NSPER+R, integrating these signals as intrinsic rewards to jointly improve replay quality and exploration. Experiments on DeepMind Control Suite tasks show that NSPER and NSPER+R improve training efficiency and convergence speed compared to existing methods in image-based RL.
Read the full abstract

Machine Learning

Pathology Transport: Optimal-Transport Explanations for Clinical Data, and When Their Heatmaps (Fail to) Localize Disease

Generative models promise a route to explainable clinical AI: rather than probe a classifier, model the distributions of healthy and diseased patients and read explanations off the geometry between them. We build such a system - an optimal-transport rectified flow trained between two clinical distributions - and use it to ask a pointed question the field too rarely tests: do the resulting explanation heatmaps actually localize disease? On tabular tumour biomarkers (Breast Cancer Wisconsin) a single flow yields per-patient counterfactuals, an unsupervised malignancy score (AUROC 0.91; 0.93 +/- 0.01 across five seeds), and a label-free attribution that agrees with a supervised classifier (r ~ 0.5) - a compact, honest interpretability engine, though it never out-predicts logistic regression. Moving to chest X-rays, we show the transport heatmap is a population-level signal, not a localiser; a reconstruction-based, identity-preserving variant does localize synthetic lesions (pointing game 0.52), yet on real RSNA radiologist boxes it collapses to chance while only supervised Grad-CAM stays above it. The central result is a synthetic-to-real gap: label-free heatmaps that look compelling on planted lesions are not evidence of real localisation. We contribute a reusable optimal-transport recipe for generative explanations and a controlled benchmark for stress-testing whether they localize.
Read the full abstract

Machine Learning

CORAM: Coherent Orthogonal Rotation for Model Merging

Merging finetuned models combines specialized capabilities without joint training or access to the original data. Most methods operate by linear arithmetic in Euclidean weight space, which cannot carry the geometry of the update. Orthogonal Model Merging (OrthoMerge) uses a single orthogonal transform for each weight matrix, but such a transform cannot change singular values. We propose CORAM, which partitions each target matrix into row slices, represents every expert slice by its singular value decomposition in the corresponding base-model SVD frame, and merges the task-specific factors on their corresponding manifolds. Because manifold averaging contracts the merged update, CORAM applies an amplification coefficient $\lambda=\kappa\hat{c}$. The scale c_hat is estimated from the expert and merged update norms and is approximately $\sqrt{N}$ for $N$ experts with comparable update magnitudes. The restoration strength kappa is selected from the dispersion of expert updates without evaluating candidate merged models. This rule remains within 0.72 points of the best swept value on all evaluated suites. CORAM also includes spread slicing to distribute highly updated rows across slices and a residual pathway for non-target layers. Across four suites covering three model families, 3B to 9B scales, and language and vision-language experts, CORAM improves over OrthoMerge by 0.25 to 1.35 points and matches or exceeds the strongest weight-space baselines.
Read the full abstract

All authors Xinyi Sui, Ziran Liu, Nam Ling, Wei Wang, Wei Jiang