跳到正文
研踪PaperTrail

论文库

检索计算机领域论文,按研究方向、作者或关键词筛选。

部分采集异常
正在读取账号与研究记录…

全站更新概况

+1680新增论文
+1333其中 AI 相关
+187已有论文修订
+59新增 AI 导读

arXiv 最近完成扫描:2026-11-18 04:28 · 全站最近入库:2026-11-17 23:35各信源进度 →

统计口径与数据新鲜度

统计 2026-11-17 09:21 — 2026-11-18 09:21(北京时间) · 点击数字查看对应简报;AI 相关包含在新增论文内。

最近完成检查:2026-11-18 04:28最近发现新增版本:2026-11-17 23:35库内论文最新更新时间:2026-11-17 02:59

每分钟刷新已保存的数据。空检查不代表采集异常;新收录可能包含历史论文。AI 导读按新增版本计数。

25,786 已收录论文21,491 AI 相关7,725 有修订112 HF 社区发现

浏览论文库

时间范围跟随排序:版本更新、本站收录或首次发表。最新收录可能包含历史论文。部分信源异常,查看状态 →

找到 2,395 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-18 04:27(北京时间)
新论文 · v1arXiv:2611.21604

FaceJev: Direct Visual Decisions for Face Forgery Detection

Siran Peng, Tianshuo Zhang, Yibo Xue, Ajian Liu, Haoyuan Zhang, Weisong Zhao, Shuhuan Chen, Xiangyu Zhu 等 9 位作者

原文摘要

Face forgery detection methods often perform well on seen manipulations but generalize poorly to unseen ones. Multimodal Large Language Models (MLLMs) can recognize facial parts and assess their visual plausibility. Existing MLLM-based methods typically express these judgments through textual explanations, but seldom treat individual part decisions as explicit learning targets. We present FaceJev, an MLLM-based method for face forgery detection through direct visual decisions. Following the Jev formulation, FaceJev answers eight independent questions about facial parts using a shared image encoding and reads anomaly probabilities from the native answer logits. A face is classified as forged if any part is predicted to be abnormal, directly linking the detection result to the part decisions. We construct PartForensics, a dataset of approximately 160k images with binary anomaly targets for the eight facial parts, to supervise these decisions. We further propose FS-GRPO, which optimizes each part decision through an exact reward comparison with its alternative, using both detection and part rewards. FaceJev outperforms the strongest baseline by 7.25 percentage points on unseen manipulations. Extensive ablation studies validate our design choices and the effectiveness of the main components.

cs.CV

新论文 · v1arXiv:2611.21559

参考强制:基于区域引导漂移校正的长视频生成

Reference Forcing: Long Video Generation with Region-Guided Drift Correction

Sheng Zhao, Haitian Zheng, Weikai Lin, Junjie Zhao, Yuhao Zhu

AI 速读 基于摘要

提出Reference Forcing方法,利用初始帧作为外观参考,通过教师蒸馏和值校正来减少长视频生成中的主体与背景漂移。

cs.CV

查看原文摘要

Long video generation should keep the subject and background consistent while supporting rich, natural motion. However, autoregressive models are often distilled with distribution matching distillation (DMD) over short windows, which does not train them to keep this consistency over long ranges. As a result, the subject and background can drift over time. Methods that manage the key-value (KV) cache retain or retrieve history to reduce drift. Yet this history may itself have drifted. We propose Reference Forcing, which uses the initial frames as an appearance reference. We first build a teacher that uses attention masks to force each region of a new frame to attend to the same region in the initial frames. The teacher finds these regions with off-the-shelf segmentation models. We then distill this behavior into a separate attention path of the student, which needs no external models at inference. We further introduce value correction, which corrects drifted values in the KV cache toward the matching values of drift-free initial frames. Experiments show that Reference Forcing achieves state-of-the-art subject and background consistency among long-video methods on different base models, and its advantage grows with video length. Ablations show both components are needed to gain consistency without losing motion. More visualizations and code are available at https://zhaosheng-thu.github.io/reference-force/

新论文 · v1arXiv:2611.21533

中国电力数据资产的激励对齐:区块链赋能的贡献信用框架

Incentive Alignment for Electricity Data Assets in China: A Blockchain-Enabled Contribution-Credit Framework

Jiazheng Zhang, Shouwei Li, Yutian Deng, Wei Wei, John Cartlidge

AI 速读 基于摘要

提出区块链赋能的贡献信用框架,利用延迟证据、动态信用、托管和报告关联惩罚促进电力市场中真实报告贡献并实现激励分配。

cs.CY · cs.DC

查看原文摘要

Modern electricity markets increasingly rely on reported operational contributions to allocate incentives, recognise cooperation, and coordinate activities across multiple organisations. However, supporting evidence for many reported contributions becomes available only after subsequent metering, dispatch, settlement, or auditing processes, making it difficult to reward genuine contributions while discouraging strategic over-reporting. Existing approaches largely address data sharing, performance verification, blockchain record management, or incentive allocation in isolation, providing limited support for integrating delayed evidence into a unified incentive mechanism. This paper proposes a blockchain-enabled contribution-credit framework for truthful reporting under delayed evidence. The framework links authenticated evidence to contribution recognition through dynamic credit updating, escrow, report-linked penalties, and automated incentive execution. Developed using the Chinese electricity market as the primary setting, it is demonstrated through two representative coordination scenarios, with adaptations illustrating its applicability to UK and US electricity-market structures. A controlled heterogeneous-agent simulation evaluates the behavioural effects of the proposed mechanism, while security analysis, public-data workflow analysis, and blockchain implementation experiments assess its practical feasibility. The results show that integrating delayed evidence with dynamic credit, escrow, and report-linked penalties promotes truthful reporting while providing a transparent and auditable process for translating verified operational contributions into incentive allocation. The proposed framework offers a practical mechanism for improving accountability, coordination, and incentive compatibility in collaborative electricity markets.

新论文 · v1arXiv:2611.21485

迷失在走廊:跨重访的持久空间记忆基准测试

Lost in the Hallway: Benchmarking Persistent Spatial Memory Across Revisits

Henry Zheng, Weibo Gong, Yang Yue, Rui Huang, Yang Zhang, Gao Huang

AI 速读 基于摘要

作者提出ReSpaceBench基准,用于评估具身智能体在多次重访中保持空间记忆的能力,发现现有模型与人类差距显著。

cs.CV

查看原文摘要

Persistent spatial understanding is essential for embodied agents to operate reliably in environments they repeatedly encounter. It requires recognizing revisited places despite viewpoint changes, distinguishing them from similar-looking rooms, and updating memories as environments change. Yet existing evaluations often leave physical place identity across separated visits implicit. We introduce ReSpaceBench, a benchmark comprising 3,244 questions across 590 synthetic and real-home scenes, organized into 9 tasks spanning room association, persistent state tracking, and downstream reasoning. Controlled walkthroughs define visits by physical room entry and deliberately include same-type room distractors and cross-session object changes. Our evaluation of 32 models and 3 agentic video-understanding frameworks reveals substantial limitations: the strongest evaluated zero-shot model achieves 50.5% overall accuracy and 52.2% on a 450-question subset, compared with 90.1% for human annotators. The evaluated agentic frameworks offer no overall improvement over direct inference with the same base model. Analysis of agent execution traces reveals evidence omissions and incorrect cross-visit room associations as major failure modes, suggesting that iterative evidence seeking alone does not ensure consistent spatial memory. Further analyses expose a gap between recognizing familiar rooms and identifying the correct visited room, while annotation-guided sampling highlights the importance of preserving intervening visits and transitions. We also provide RS-30K, a separate training dataset from synthetic sources with no template, scene, route, or video overlap with the benchmark. Fine-tuning Qwen3-VL-8B raises overall accuracy from 35.5% to 60.4%, with the largest gains in association and improvements extending to real-life scenes.

新论文 · v1arXiv:2611.21478

Jetriever:将智能体记忆检索从相似度转向效用

Jetriever: Shifting Agent Memory Retrieval from Similarity to Utility

Ruizhe Li, Licheng Zhang, Mingxuan Du, Zhendong Mao

AI 速读 基于摘要

Jetriever 用 Jev 决策分数对全部记忆池按“对当前问题的有用性”排序,取代相似度检索,并在 InMind 与 MemoryAgentBench 上报告了优于对照的排序与得分表现。

cs.CL

查看原文摘要

A useful memory need not resemble the question it helps answer. Jetriever ranks each stored memory by Jev's decision score for its usefulness to the current question, replacing similarity search with full-pool utility scoring. On InMind's 125 indirect-query tasks, it places the gold exchange first in 84.8% of cases; matched BM25 and MiniLM controls reach 1.6% and 0.0% when indexing user turns. A 20-question comparison separates finding evidence from ranking it: Jev and Qwen3.5 9B each find 90% of gold exchanges by rank 50, while Laya finds 30%; at rank one, Jev reaches 85% and Qwen3.5 9B 40%. On a dimension-proportional random sample of 60 MemoryAgentBench questions, Jetriever's four-dimension score is 55.4%, against 44.3% for ReFind on the same questions. Jev retrieval costs USD 0.00834 per InMind question and USD 0.03103 per MAB question on average. These results show the value of world knowledge, scores that rank decisive memories first, and affordable full-pool scoring.

版本更新 · v2arXiv:2611.16591

Singularity-Bench: Can Agents Build Agents?

Yang Zhou, Krish Agarwal, Zhuoming Chen, Zerui Wang, Zhenyu Zhang, Wenshuang Li, Xinpeng Wei, Zhichen Zeng 等 19 位作者

原文摘要

Frontier AI agents increasingly succeed on realistic, long-horizon tasks and are beginning to improve the systems through which they operate. Yet existing benchmarks often restrict agents to narrow action spaces, use simplified environments, or rely on verification methods that are unreliable for evaluating arbitrary agent-generated systems. We introduce Singularity-Bench, an open-ended benchmark for autonomous AI development with six end-to-end tasks spanning pre-training, mid-training, post-training, inference, the agent harness, and cluster orchestration. Each task requires agents to implement core systems from a minimal starting point and evaluates them against mature human-engineered baselines on identical hardware. Additionally, through extensive red-teaming covering 9 of the 10 categories in the OWASP Top 10 for Agentic Applications, we expose reward-hacking vulnerabilities in the conventional Harbor framework and develop Breakwater, an extension of Harbor that blocks all attacks tested in our study. Across 11 agents built from 7 frontier models and 2 harnesses, Opus 5.5 with Claude Code matches or outperforms the human baseline on all six tasks with a Singularity-Index of 1.25, suggesting it may have the skills needed to reach the singularity. The remaining agents average only 0.60. Our vision is for Singularity-Bench to track how much of AI development agents can already automate and when they may transition from assisting to taking end-to-end responsibility for building stronger successors.

cs.AI

版本更新 · v2arXiv:2608.21847

CALE: Color-Adaptive Low-Light Enhancement via Learnable Intensity Coordinates

Yi Ai, Zheng Chen, Yuanhao Cai, Yulun Zhang, Xiaokang Yang

原文摘要

Small brightness differences in dark regions are difficult to recover, and their spacing in the coordinates used for restoration depends on the intensity mapping. A globally learned Box--Cox exponent controls the full response, coupling different brightness intervals and applying the same mapping to every image. We propose Color-Adaptive Low-Light Enhancement (CALE), which combines an image-conditioned intensity representation with structure-conditioned rectified flow (\mbox{SR-RF}). The representation, Box--Cox Intensity Adaptation (BIA), refines a normalized Box--Cox transform using a monotone piecewise-linear curve. Compact statistics of the low-light observation determine the output widths of brightness intervals, allowing the response to vary across images. Positive, normalized slopes preserve intensity order and endpoints and give the intensity mapping an analytic inverse. Combined with polar chromaticity, BIA defines the coordinates for the \mbox{SR-RF} network. The mapping and restoration network are trained jointly so that the reconstruction objective shapes the intensity response. Experiments on paired benchmarks and unseen datasets show competitive restoration performance across diverse low-light scenes, with qualitative comparisons illustrating the recovery of scene colors, contrast, and detail. Coordinate ablations separate the effects of interval flexibility and input conditioning, while Full and Tiny models offer quality--computation trade-offs.

cs.CV

新论文 · v1arXiv:2611.21402

Harmonic-Mean Aggregation of Continuous-Time Markov Chains with Reachability Error Bounds

Bingqing Hu, Landon Taylor, Zhen Zhang

原文摘要

Abstracting a continuous-time Markov chain by grouping its states into blocks raises a fundamental question: which rates should the abstract chain use when states within a block have different exit rates and transition probabilities? We study harmonic-mean aggregation, which matches the mean waiting time and next-block distribution of a local experiment that uniformly selects a state within the block and observes its next jump. Equivalently, the abstract transition rates are averages of the concrete rates with inverse-exit-rate weights. Since the aggregation does not preserve reachability probabilities in general, we derive theoretical error bounds for time-bounded reachability. We decompose the error into a blockwise mean and centered-within-block deviations, and show that the rate mismatch has zero weighted mean. Exploiting this cancellation and differences between abstract reachability values, we obtain bounds valid at every time horizon through a linear comparison system with two quantities per block, without computing concrete reachability values. All results are formalized and proven in Lean 4.

cs.LO

新论文 · v1arXiv:2611.21392

VideoNexus: Video Perception Beyond the Reasoning Context

Yancheng Long, Xuanyu Zheng, Wei Chen, Luoyuan Zhang, Haoyang Xu, Bin Wen, Tingting Gao

原文摘要

Understanding long videos requires locating evidence, resolving visual details, and integrating observations into an answer. Even when relevant evidence is present, performing these operations within long, dense visual sequences burdens both perception and reasoning. Video agents enable selective inspection, but returning frames to the planner still accumulates visual information in its reasoning context. Delegating perception reduces this burden, yet sparse sampling can miss decisive details, and per-frame descriptions can preserve context growth. We introduce VideoNexus, a training-free framework that separates detailed video perception from evidence-driven planning and reasoning. The planner controls where to inspect, what to resolve, and how densely to sample. Perception sub-agents process native audio-video in independent contexts and return compact, temporally grounded evidence for iterative search and verification. Using Gemini 3.8 Flash for both planning and perception, VideoNexus achieves 88.6% accuracy on full LVBench, exceeding same-backbone static ingestion and built-in agentic processing by 3.6 and 1.1 points, respectively, at 44% of the latter's measured API cost. A paired LVBench subset ablation shows lower planner input and higher accuracy. With Kimi K3 as the shared backbone, VideoNexus improves over the official Video-GAIA ReAct harness by 8.5 points in our evaluation. These results support compact evidence exchange for combining detailed observation with efficient planning and reasoning.

cs.CV

新论文 · v1arXiv:2611.21339

MovieRoll: A Systematic Framework for Rolling Out Long-Form Videos

Yunyang Ge, Weijie Kong, Ryan Xu, Yue Wu, Qi Tian, Jiangfeng Xiong, Bin Lin, Mingzhe Zheng 等 12 位作者

原文摘要

Streaming video generation must sustain quality beyond the training horizon while handling self-generated history and growing cache costs. To address these challenges, we present MovieRoll, a base-to-distilled framework that connects streaming fine-tuning, long-rollout distillation, and long-horizon inference. We first establish a robust multi-step base through Window-wise Teacher Forcing (WinTF), which jointly denoises chunks at progressive noise levels with bidirectional attention within each window and causal access to preceding history. To retain historical context efficiently, hierarchical KV-cache compression (HieraKV) exploits layer-wise attention patterns, retaining full history in shallow layers and sink--recent history in deeper layers. This reduces cache memory by up to 78% while improving generation quality. Building on this base, cumulative long tuning with distribution matching distillation (CuLT-DMD) holds parameters fixed throughout each long self-rollout and averages local gradients before updating. Each update therefore combines supervision across temporal positions while retaining only one local computation graph, enabling efficient long-rollout training of a few-step generator. To extend generation beyond the training horizon, KV recomputation bounds context for both models, while guidance annealing (AnnealCFG) reduces guidance toward the clean endpoint to stabilize MovieRoll-Base. Complementing these modeling advances, FBCA attention and PS-VAE decoding support efficient training and streaming output. On HeliosBench, MovieRoll-Base remains competitive at 30s after training on videos of up to 15s, and MovieRoll-Distilled achieves the highest total scores among the compared methods across single- and multi-event evaluations up to 60s, combining sustained naturalness with active motion.

cs.CV

版本更新 · v2arXiv:2611.09387

A Bridge from Paste to Place: Physics-Guided One-Step Image Composition via Object-Scene Interaction

Hongzhen Li, Youwei Pan, Yingfang Zhu, Sen Zhu, Fengjie Zhu, Leilei Cao

原文摘要

Realistic image composition makes a pasted object appear naturally present in a scene while preserving the pasted layout and fine foreground and background details. Reference-based insertion methods preserve overall object identity but can alter pose and fine details, often at substantial computational cost. Foreground harmonization and shadow generation address individual aspects of this task, but neither alone ensures natural integration. An efficient, unified solution remains an open challenge, and large-scale, high-quality datasets and benchmarks for this task are lacking. To address these challenges, we introduce Paste2Place, an end-to-end, physics-guided flow-based model that learns a direct latent bridge from pasted inputs to realistic composites, guided by image-space corrections analytically derived from predicted physical components of object and scene appearance changes. To support training and evaluation, we introduce P2P-Data with 86K manually reviewed synthetic and real-scene training pairs, including dense physical annotations from our Blender rendering and validation pipeline for the synthetic subset. P2P-Bench combines paired compositions and manually pasted inputs from collected images into 1.5K examples for evaluating composition under user-chosen placements. Extensive qualitative and quantitative experiments demonstrate that Paste2Place achieves substantial advantages in image composition quality and inference speed across diverse scenarios.

cs.CV

版本更新 · v2arXiv:2404.00551

Gaussian-Channel Regularity and Distribution Learning with Flow Matching

Yuan Gao, Jian Huang, Yuling Jiao, Shurong Zheng

原文摘要

We study distribution learning by flow matching with linear interpolation from a Gaussian source under Gaussian-channel regularity, a condition on the target distribution. For an integer $r\geq2$, we establish spatial regularity of order $r$ for the population velocity, together with temporal Lipschitz continuity on the full interval and local time analyticity before the terminal time. We use these properties to construct a finite multiscale ReLU sieve with a spatial Lipschitz bound independent of the truncation radius. For an exact minimizer of the empirical flow matching risk over this sieve, we obtain an upper bound on the expected Wasserstein-$2$ error of order $n^{-r/(d+2r)}$, up to a logarithmic factor, where $n$ is the sample size and $d$ is the dimension. The bound includes Euler discretization up to time one, without early stopping. The conditions hold at $r=2$ for targets with strongly convex $C^2$ potentials and two-sided Hessian bounds. They also hold at every fixed order $r\geq2$ for distributions obtained by convolving a law of bounded support with a nondegenerate isotropic Gaussian. For sufficiently large fixed constants defining the target class, we prove a minimax lower bound of order $\max\{n^{-1/2},n^{-(r+1)/(d+2r)}\}$.

stat.ML · cs.LG

新论文 · v1arXiv:2611.21274

SURMA: Efficient Fine-Grained Communication Offload for SuperPOD

Haoran Shang, Weifeng Zhang, Mengzhu Li, Huizheng Wang, Sihan Guan, Xinyang Huang, Bowen Liu, Ran Shu 等 17 位作者

原文摘要

GPU SuperPODs underpin large language model (LLM) serving, but growing bandwidth and round-trip times make inter-GPU communication consume increasing compute resources. SM-driven remote memory access occupies shared memory or registers for in-flight data and competes for SM-to-NoC bandwidth. Dedicated engines can free these resources, but must handle fine-grained transfers, down to a 576\,B KV-cache entry. NVIDIA's CopyEngine uses less than 1\% of NVLink bandwidth at 7\,KB granularity. We present SURMA, an on-die Scale-Up Remote Memory Access engine. Adapting RDMA requires overcoming slow doorbell submission, excessive per-application connection state, head-of-line blocking of completion notifications on shared connections, and costly ordering for multipath transfers. SURMA uses a credit-based shared send queue to reduce submission latency from 1.2\,$μ$s to 400\,ns; fused write-with-atomic semantics to share connections while independently notifying each transaction; and Network Per Path Ordering to ensure data arrives before the completion flag without Sender or Receiver Ordering. We implement SURMA in RTL on a two-FPGA prototype and evaluate it with two 8-GPU NVLink servers. At 512\,B granularity, SURMA saturates attainable bandwidth with a single SM, whereas NVLink needs over 14\% of the SMs. SURMA occupies under 0.5\% of GPU area and accelerates inference on four models by 1.05--1.26$\times$ over NVLink.

cs.AR

新论文 · v1arXiv:2611.21253

QuantileGS: Quantile-Based Conformal Bounds on RGB Rendering Errors for Novel View Synthesis with 3D Gaussian Splatting

Haiyang Xu, Ziye Zhang, Bin Pan, Xia Xu, Zhenwei Shi

原文摘要

Uncertainty quantification for novel view synthesis with 3D Gaussian Splatting (3DGS) assesses rendering reliability at unseen viewpoints. This reliability depends on the discrepancy between rendered and true colors, which cannot be measured without ground-truth images. We therefore quantify uncertainty in this unknown discrepancy through upper bounds on per-pixel RGB mean absolute error. However, learned bounds may underestimate these errors and overstate rendering reliability, motivating calibration to a prescribed coverage level. In this paper, we propose QuantileGS, a post-hoc framework that combines learned rendering-error quantiles with scale-normalized conformal calibration to produce spatially adaptive error bounds. The framework comprises three phases. Phase 1, Rendering-Error Computation, measures RGB mean absolute error between observed images and predictions from a frozen 3DGS model. Phase 2, Ordered Gaussian Quantile Fields, learns two scalar attributes per Gaussian to predict lower and upper rendering-error quantiles. Phase 3, Scale-Normalized Conformal Calibration, calibrates the upper prediction on held-out images, using the quantile gap to scale the correction. Under the stated assumptions, the resulting bounds provide finite-sample marginal coverage. At a nominal coverage level of 90%, calibration raises mean empirical test coverage from 69.6% to 89.6% on LLFF and from 80.1% to 92.1% on Mip-NeRF 360. On LF, coverage remains conservative at 96.3% before and 96.4% after calibration. Applications to active view selection, Gaussian pruning, and RGB-D simultaneous localization and mapping illustrate how these bounds inform downstream decisions.

cs.CV

版本更新 · v2arXiv:2611.17060

Ropedia Xperience 2: A Gold Standard for Scaling Physical AI Environments and Behaviors

Hao Li, Fangzhou Hong, Zhaoxi Chen, Hui En Pang, Kairui Hu, Mingze Sun, Runmao Yao, Xinhao Ji 等 22 位作者

原文摘要

Large foundation models such as GPT-6 Astra and Claude Fable 5.1 have recently been shown to control robots directly from visual observations. They nevertheless remain brittle on long-horizon, contact-rich tasks in real-world environments. Further progress depends on two complementary resources: demonstrations of human behavior that supervise how tasks are performed, and interactive environments in which policies can be rolled out, evaluated, and improved in closed loop. Language models have advanced along well-characterized scaling laws; the corresponding question for physical AI is how capability scales with environments and behaviors, which first requires identifying which data to scale and how to produce and reuse such data continuously. To address this question, we introduce Ropedia Xperience 2, comprising 1,589 released metric-scale 3D scans and a collection of 4,230 egocentric recordings of 74 tasks (131 hours). We further show that each capture can be repurposed for robot learning: scenes are converted into physically interactive simulation environments, human motion is retargeted to whole-body robot trajectories, and additional robot behaviors are obtained through virtual teleoperation and automated long-horizon trajectory generation. We propose this joint capture of environments and behaviors as a gold standard for physical AI data collection, laying a data foundation for establishing scaling laws over environments and behaviors.

cs.CV

版本更新 · v6arXiv:2605.18267

SRC-Flow: Compact Semantic Representations Enable Normalizing Flows for Image Generation

Longtao Jiang, Jianmin Bao, Zhendong Wang, Xin Tao, Zhihui Li, Xiaojun Chang

原文摘要

Normalizing flows (NFs) provide exact likelihoods and deterministic invertible sampling, but have historically lagged behind diffusion models for large-scale image generation. We identify a key obstacle: NFs are required to learn a single invertible transport over the full ambient space, making them highly sensitive to high-dimensional representations. This leads to a semantic-capacity mismatch in modern visual representation spaces, where semantic information is compact but encoded in overcomplete features. We propose SRC-Flow, which introduces a Semantic Representation Compressor (SRC) to compact high-dimensional RAE features into a low-dimensional semantic space before flow modeling and preserve reconstruction through the frozen RAE decoder. This compact space reduces the modeling burden of NFs and enables effective likelihood-based generation in semantic representation space. We further adopt constant noise regularization tailored to the fixed unconditional bijection learned by flows. On ImageNet 256x256 and 512x512, SRC-Flow achieves state-of-the-art generation quality among normalizing flow methods, with gFID scores of 1.65 and 2.07 under classifier-free guidance, while retaining exact likelihood computation in the compact semantic representation space and deterministic invertible sampling at the flow level. Codes and models will be released in the future.

cs.CV

新论文 · v1arXiv:2611.21152

From Repositories to Training Environments: Scaling Long-Horizon Coding Agents with Environment Evolution Loop

Yang Tian, Tianyu Zheng, Xiyuan Yang, Yaxin Du, Chengran Yang, Bo Zhao, Jian Yang, Bryan Dai

原文摘要

Training coding agents for long-horizon tasks requires environments with verifiable feedback. Existing pipelines often follow a linear process that ends with validation, establishing usability but providing little guidance on improving training value. Recursive self-improvement (RSI) often requires model rollouts or model training followed by downstream evaluation, with costs and feedback delays amplified on long-horizon tasks. How can we use timely, reliable proxy signals to improve training value and synthesize environments at scale? We present RepoFactory, an automated pipeline that converts real repositories into verifiable environments where agents build complete projects from scratch given a language specification. Its Environment Evolution Loop optimizes a multiplicative objective combining validity, structural coupling, and implementation scope through proxy feedback. RepoFactory produces 10,159 environments across 17 software domains. Training Qwen3.5-35B-A3B on these environments yields RepoHorizen-35B-A3B, which achieves gains of 22.6 percentage points on NL2Repo-Bench and 8.5 points on average across five benchmarks covering long-horizon coding and agentic tasks in real-world production settings, compared with 17.5 and 6.4 points without environment evolution. These results highlight the training value of RepoFactory environments and the additional gains from the Environment Evolution Loop.

cs.CL

新论文 · v1arXiv:2611.21114

VAE Rhythm: Temporal Decoding Traces for AI-Generated Video Detection

Ke Sun, Shuwei Li, Haiyang Mei, Mike Zheng Shou

原文摘要

Visible artifacts are an unreliable cue for detecting generated videos, because newer generators learn to remove them. We instead detect a trace left by the generation pipeline itself. Most video generators decode latents with a 3D causal VAE that groups consecutive frames in time, so the fidelity of generated frames rises and falls periodically with their position in a group. We call this periodic trace VAE Rhythm. Real videos are captured frame by frame and do not have it. To reveal it without training, we propose a phase scan. The scan reconstructs a clip with an off-the-shelf VAE from successive start offsets and measures the periodicity of the reconstruction error across these offsets. We find VAE Rhythm in most of the 39 variants of 31 generators we test, including the closed-source Sora, Kling 3.0 and Seedance 2.0. We use the rhythm as a zero-shot score, as a plug-in for existing detectors and as a data engine that pairs real videos with their VAE reconstructions to train our detector CaDet. On DF26, the training-free rhythm score outperforms all 15 retrained baselines, and as a plug-in it improves all 19 tested detectors by 6.7% to 23.7% AUROC. CaDet achieves the best AUROC on AIGVDBench, DF26 and three 2026 closed-source models, and its variant trained without any generated video still exceeds 90% on AIGVDBench.

cs.CV

新论文 · v1arXiv:2611.21105

CivilBench: Benchmarking Vision-Language Models for Multimodal Information Integration in Civil Engineering

Dan Yuan, Yue Wen, Zhili He, Zhenkui Zhou, Shicen Liu, Ziqiang Zheng

原文摘要

Artificial intelligence (AI), particularly large language models (LLMs) and vision-language models (VLMs), is increasingly applied in civil and structural engineering for disaster assessment, structural analysis, and engineering decision support. However, the extent to which current VLMs possess the reasoning capabilities required for reliable engineering judgments remains unclear. In this work, we introduce CivilBench, a comprehensive benchmark for evaluating VLMs in civil engineering and disaster-related scenarios. CivilBench covers three major task families: Sequential Reasoning, Image-based Structural Understanding, and Open-world Understanding, encompassing numerical reasoning, structural damage assessment, risk inference, causal analysis, and repair planning. The benchmark is constructed through a systematic pipeline of domain-specific data collection, engineering-oriented captioning, expert annotation, question design, and quality verification. Evaluation of representative state-of-the-art VLMs reveals substantial limitations in domain-specific reasoning, with Sequential Reasoning emerging as a major bottleneck, particularly when models must infer numerical relationships from sparse structured observations. Further experiments show that numerical priors and external knowledge can alleviate these limitations, although their effectiveness depends on the model's ability to reliably interpret and use the available evidence. CivilBench provides a capability-oriented testbed for identifying reasoning limitations and evaluating progress toward reliable AI-assisted engineering decision-making. The benchmark will be publicly available at https://huggingface.co/datasets/traceuryuan/CivilBench.

cs.AI · cs.CV

新论文 · v1arXiv:2611.21099

ForceHarness: Turning Force into Action and Physics for Frozen Frontier Models

Ke Sun, Kevin Yuchen Ma, Yike Zhang, Yanzhe Chen, Mike Zheng Shou

原文摘要

Frontier multimodal models can now control robots zero-shot from images, yet many manipulation tasks depend on contact forces that images cannot show. Bringing force to an API-only frozen model raises two challenges. First, contact evolves much faster than an API model can decide, so the system must close the loop on force beneath the model. Second, a frozen model cannot interpret raw force streams, so force must be returned in a form the model can reason about. Therefore, we present ForceHarness, a harness that lets frozen models use force without demonstrations or finetuning. After calibrating hand force from a wrist sensor or joint torques alone, ForceHarness addresses the first challenge with self-terminating force primitives that regulate force on their own and return at a contact event. To address the second challenge, a physics translation layer turns each force trace into physical quantities with uncertainty as the tool result. To evaluate ForceHarness against expert force controllers and learned policies under identical conditions, we also build HiddenPhys, a seven-task MuJoCo benchmark with hidden physical variables. On a real arm without a wrist force sensor, ForceHarness raises the zero-shot success rate of GPT-6 Astra from 3.3% to 90.0%. In simulation, it adds 17.5 percentage points on the public ForceBench suite and 34.3 percentage points on HiddenPhys, surpassing force-aware policies trained on demonstrations on both benchmarks and matching hand-written expert controllers on HiddenPhys. Ablations further show that smaller models make significantly better decisions when given physical quantities rather than raw force values.

cs.RO

新论文 · v1arXiv:2611.21072

Sage S2.1 Flash Technical Report

Kun Cai, Haodong Chen, Guangran Cheng, Qingxiu Cheng, Xu Dai, Guanting Dong, Qiming Ge, Lixin Gu 等 206 位作者

原文摘要

We introduce Sage S1.5 Flash, a 320-billion-parameter general-purpose mixture-of-experts model designed to reason over scientific evidence and act in interactive environments. Sage Sci-Perception uses dedicated encoders for molecular graphs and biological sequences and represents spectra as ordered patches of measurements. Modality-specific projectors map these representations into the input space of a shared backbone, which processes them alongside text and images. To develop the model's agent capabilities, we construct verifiable environments for tool use, search, and coding, as well as specialized workflows involving machine learning experiments and scientific software. Task completion is checked against execution results, changes in environment state, or submitted artifacts. We evaluate Sage S1.5 Flash on 33 benchmarks across seven domains. In the primary comparison against five baselines, it ranks first on OfficeQA Pro (72.2), MLE-bench Lite (90.3 HumanRank), Harvey LAB, and SpreadsheetBench~2, and second on OSWorld-Verified (85.2), WebArena-Verified (67.6), and DeepResearch Bench~II (54.2). These results show particular strengths in using documentary evidence, producing verifiable workbooks, and carrying out machine learning tasks, alongside competitive performance in desktop and browser workflows that require the model to act on visual feedback.

cs.AI

新论文 · v1arXiv:2611.21060

MetricHand: Geometry-Aware Flow Matching for 4D Hand Reconstruction in Metric World Space

Yibo Bai, Likui Zhang, Jinglei Zhang, Zhengbin Pang, Xingshen Song, Xiaoqing Yin, Jinsheng Deng

原文摘要

Egocentric hand motion provides physical motion supervision for dexterous manipulation and vision-language-action pretraining. Reliable use of this supervision requires accurate and temporally coherent 4D hand reconstruction in metric world space. Existing methods typically lift framewise estimates produced under weak perspective into world coordinates. By ignoring joint-dependent depth variation at close range, this approximation couples errors in translation, orientation, and articulation that persist after world lifting and worsen under occlusion. We present MetricHand, a geometry-aware flow matching framework for 4D hand reconstruction in metric world space. MetricHand encodes depth priors from hand-object interaction regions into joint-associated representations that preserve local spatial structure and metric scale. Fused with visual features, this geometry guides coarse metric estimation and conditions every step of temporal generation. The model jointly recovers world-space translation, global orientation, MANO articulation, and shape, while structural priors and temporal context support estimation under partial occlusion. MetricHand reduces camera-space MPJPE by 49.3% and 52.9%, absolute world-space MPJPE by 28.7% and 21.7%, and wrist localization error by 48.5% and 37.9% relative to HaWoR, respectively. World-space acceleration error decreases by 44.2% and 64.3%, demonstrating improved temporal fidelity alongside more accurate absolute localization.

cs.CV

版本更新 · v2arXiv:2611.11995

Infrathon: Benchmarking Language Agents for Real-World, Long-Horizon AI Infra Optimization

Zhengyu Hu, Zheyuan Xiao, Yike Wang, Hamish Ivison, Haoyu Wang, Yige Yuan, Rulin Shao, Yuetai Li 等 21 位作者

原文摘要

In this work, we study whether frontier agents can continually improve real-world AI infrastructure over long horizons. We introduce Infrathon, a benchmark of diverse, long-horizon AI infrastructure optimization tasks across different levels. We evaluate six models across 150 runs with 24-hour budgets, totaling over 10,000 GPU-hours. Our results show that model rankings vary with the evaluation horizon, while 135 of the 150 runs achieve further improvements during the second half of their budgets. These findings indicate that shorter evaluations can overlook both optimization gains and changes in relative model performance. We also find agents can discover solutions that outperform human-designed baselines: Claude Opus 5.5 increases the mean KV-cache compression ratio from $6.82\times$ with the human-designed Quant-VideoGen codec to $19.86\times$. Analysis of agent trajectories identifies task violations, unreliable testing, and premature completion as barriers to sustained improvement. Together, our Infrathon framework provides resources for constructing long-horizon tasks, evaluating frontier agents, and discovering good infrastructure solutions. Code is available at https://github.com/InfrathonX/Infrathon

cs.AI · cs.CL

新论文 · v1arXiv:2611.21019

MemDoctor: Stage-wise Error Attribution and Repair for MLLM Agent Memory

Ruiyang Jiang, Xinyang Liao, Weikai Huang, Yuqi Niu, Junjie Zhou, Yubin Zheng

原文摘要

Multimodal large language models (MLLMs) now assist users across sessions of images and text, carried by an external memory of what was said and shown. When such an assistant answers wrongly, the failure usually began earlier, in how a record was written, updated or retrieved, and the answer alone does not say where. We formalize multimodal memory as four stages, Write, Update, Retrieve and Answer, and define nine error types across them. A stage-resolved analysis on Mem-Gallery traces most failures to a source stage, and most of those lie before the answer. Injecting known errors against fact-preserving control edits, we then follow each tainted record downstream. Early errors survive to the answer as apparent evidence, and hurt most when no other record holds the fact. A diagnosis that trusts the quoted evidence therefore blames the stage where the error surfaces, not the one where it began. We propose MemDoctor, in which the MLLM that built the memory diagnoses its own failures. It traces each needed fact backward through the four stages, accepts a diagnosis only when the cited evidence is found in memory, and repairs that stage before answering again. On a 4B backbone, MemDoctor improves accuracy on both Mem-Gallery and MemLens, outperforming all evaluated 8B baselines, and the same loop also improves MuRAG and A-Mem.

cs.CL

版本更新 · v4arXiv:2602.05302

PieArena: Ranking and Profiling Language Agents in Realistic Negotiation Scenarios

Chris Zhu, Sasha Cui, Will Sanok Dufallo, Runzhi Jin, Zhen Xu, Linjun Zhang, Daylian M. Cain

原文摘要

We present an in-depth evaluation of LLMs' ability to negotiate, a central business task that requires strategic reasoning and theory of mind to create and claim economic value. To do so, we introduce PieArena, a large-scale negotiation benchmark grounded in multi-agent interactions over realistic scenarios adapted from MBA negotiation courses at an elite business school. We evaluate language agents across three pairing regimes: mirror-play, cross-play, and human-LM play. We develop a ranking model for continuous negotiation payoffs that yields order-invariant, uncertainty-quantified leaderboards while correcting for systematic experimental asymmetries. We further study the effects of shared-intentionality agentic scaffolding and find asymmetric gains, with large improvements for mid- and lower-tier LMs and diminishing returns for frontier LMs. As calibration anchors, we collect human-human and human-LM negotiation data from trained business school students, finding that a representative frontier language agent matches or exceeds this human baseline in our evaluation settings. Beyond deal outcomes, PieArena provides a multi-dimensional behavioral profile that reveals cross-model heterogeneity in instruction compliance, computation accuracy, as well as judge-assessed deception and reputation, illustrating the value of evaluation beyond outcome-only leaderboards.

cs.AI

新论文 · v1arXiv:2611.21003

K2-Horizon: A Frontier Fleet of Open Source Language Models

K2 Team, Zhengzhong Liu, Haonan Li, Xuezhe Ma, Alexander Moreno, Subham Sekhar Sahoo, Rupesh Kumar Srivastava, Mikhail Yurochkin 等 95 位作者

原文摘要

We present K2-Horizon, a fleet of six fully open language models, from a 0.9B dense model to a 375B-A23B MoE, built for general-purpose reasoning and agentic work across math and science, software engineering, deep research, office automation, and beyond. The models lead or are competitive with strong open models of their size on agentic and coding benchmarks. On the modeling side, to strengthen reasoning, we introduce Mixture-of-Value Attention (MoVA), applying MoE to attention value projections, and softplus attention gating, a more flexible, theoretically motivated gate. Our 36B-A4B model uses both and nearly matches a 32B dense model trained on the same data, with one-seventh the active parameters. We also introduce Uno, lossless diffusion adapters that speed up decoding by up to 2.2x without a separate draft model. Our release follows the LLM360 definition of open source for transparency and reproducibility: the full development history, including intermediate checkpoints from every training stage, pre-training data, training logs, and all training, evaluation, and infrastructure code, notably our production-scale xLLM framework for dense and MoE training and the RL360 reinforcement-learning stack. Beyond the data itself, we release and document in detail every pipeline that built it, the part of model development that open releases most rarely disclose. These pipelines produced roughly 10T synthetic pre-training tokens, showing that synthetic data can be a major ingredient of pre-training at scale. Four of the models share one 22T-token pre-training corpus and recipe, revealing consistent scaling behavior and enabling controlled studies of scale and sparsity. All artifacts are available at https://ifm.ai/k2/

cs.AI

新论文 · v1arXiv:2611.20992

MedJev: A System One Vision-Language Decision Model for Fast and Effective Medical Image Understanding

Fan Zhang, Hao Wu, Jinyang Wu, Hongyu Qu, Zheng Lian, Dongfu Yin, Qingsong Wen, Pheng-Ann Heng

原文摘要

Medical image understanding with vision-language models (VLMs) typically relies on autoregressively generating reasoning and answer tokens. These unnecessary tokens make the inference cost scale with the response length and force the answer to be parsed from free-form text, incurring avoidable latency and decision errors. To this end, we introduce \textsc{MedJev}, a \emph{system one} vision-language decision model that maps medical images, questions, and candidate answers directly to decisions in a single multimodal forward pass, without any autoregressive generation process. For fast inference, \textsc{MedJev} removes the upper language layers and the generative output pathway, while keeping the full vision pathway and all visual tokens. On this shortened backbone, a decision head scores each candidate by pretrained query--key matching and refines the scores with a lightweight multilayer perceptron (MLP) residual, coupling the pretrained matching structure with task-specific nonlinear discrimination. Built on Qwen3.5-4B, \textsc{MedJev}-4B improves both accuracy and speed over the original model on each of four medical image classification and visual question answering benchmarks, with an absolute accuracy gain of \textbf{7.93\%} and a \textbf{6.04$\times$} end-to-end speedup. Further comparison with 20 cutting-edge models shows that \textsc{MedJev}-4B achieves state-of-the-art performance and inference speed among 4B--9B general and medical VLMs, indicating that direct multimodal decision making is an effective route to fast and effective medical image understanding.

cs.CV

新论文 · v1arXiv:2611.20991

FH-Sampling: Frequency-aware Hypergraph-driven Oversampling for Imbalanced Time-Series Classification

Yunrong Hao, Ruohao Dai, Zhen Liu, Guan Wang, Wenbin Pei, Qiang Zhang

原文摘要

Imbalanced time-series data appears commonly in diverse applications, such as medical diagnosis. When trained on imbalanced data, classifiers tend to exhibit a bias toward the majority class, making it difficult to identify minority-class instances accurately, despite their critical importance. Sampling techniques have proven effective in rebalancing imbalanced data. However, existing interpolation-based oversampling methods typically rely on pairwise interpolation, limiting their ability to capture high-order inter-sample relationships, and their inherent smoothing effects tend to suppress high-frequency components, compromising spectral fidelity. Therefore, we propose a Frequency-aware Hypergraph-driven Oversampling method (FH-Sampling) for imbalanced time-series classification. In FH-Sampling, a hypergraph is first constructed in the original time-series space to capture complex high-order inter-sample relationships, and sampling ratios are adaptively assigned to mitigate the impact of noisy samples. The discrete wavelet transform decomposes each sample into frequency components. Guided by the sample hypergraph, FH-Sampling combines high-frequency components from related samples while retaining the target sample's low-frequency component to help preserve important frequency characteristics. Experiments on 24 imbalanced datasets demonstrate the advantages of FH-Sampling for imbalanced time-series classification. We further apply FH-Sampling to aircraft trajectory generation. Experiments on simulated aircraft trajectories show that FH-Sampling produces trajectories that better match the real distribution, highlighting its potential for practical applications.

cs.LG

新论文 · v1arXiv:2611.20986

Calpha: Toward Native Biomolecular Intelligence in a General-Purpose Language Model

Yong Liu, Zhanpeng Shi, Peicong Lin, Zhenyi Zhong, Yizhou Dang, Wendi Wu, Haoling Li, Shuangjia Zheng

原文摘要

Biomolecular foundation models predict the structures of proteins and their complexes accurately, and large language models (LLMs) reason in natural language, but the two remain separate: the abilities of the former are not part of a general-purpose LLM, and the reasoning of the latter does not enter a structure prediction. Calpha aims at one general model that folds proteins and complexes, understands protein sequence and structure, and lets its language reasoning influence the structures it predicts. This report covers the first step, in which an LLM outputs single-chain protein structures. Qwen3.5-9B reads a sequence enriched with ESM2 residue features. It can respond in text or emit a token that invokes a pretrained SimpleFold structure head, which receives information about the protein only through Qwen's residue states. Under a fixed folding request, the head achieves mean TM-scores of 0.808 on the FoldBench-334 development set, 0.635 on CASP16 domains and 0.900 on a CAMEO-style PDB set. We compare these predictions with a baseline in which ESM2-3B encodes the sequence and a fine-tuned SimpleFold-3B head predicts the coordinates. The two systems also differ in trainable weights, training duration and crop length. For chains of 100--299 residues, all three paired confidence intervals include zero. At 300--600 residues, the mean differences are -0.044, -0.160 and -0.025, respectively. In a separate test, four English request types triggered folding on every evaluated example. Protein training reduced accuracy on a 350-item MMLU-Pro subset from 77.4% to 70.3%. An LLM can therefore output single-chain structures from its own states, with lower long-chain scores in the current comparisons and reduced accuracy on the evaluated MMLU-Pro subset. Complex prediction, structure understanding and the effect of language reasoning on the predicted coordinates remain future work.

q-bio.QM · cs.AI

版本更新 · v2arXiv:2605.21543

Provable Joint Decontamination for Benchmarking Multiple Large Language Models

Zhenlong Liu, Hao Zeng, Hongxin Wei

原文摘要

Benchmark data contamination has become a central challenge in LLM evaluation: when evaluation examples appear in the training data of one or more audited models, reported performance can be inflated and cross-model comparisons become unreliable. A broad line of training-data detection work designs scores to quantify how strongly a model memorizes a given data point, but these score-based methods lack theoretical guarantees. Recent conformal approaches provide provable false-identification control for a single model; however, applying them separately to each model can produce model-specific benchmarks, undermining fair comparison across models. In this work, we formalize multi-model benchmark decontamination as a joint selection problem and propose Joint Envelope Conformal Selection (JECS), a conformal procedure that enables global contamination rate (GCR) control under stated assumptions. Specifically, JECS computes per-model conformal p-values, aggregates them by the per-item maximum, and reconstructs a conservative envelope of the max-p null distribution from right-tail observations above a data-driven threshold. By applying the adaptive Benjamini-Hochberg (BH) procedure to the envelope-rescaled values, we select a benchmark with provable GCR control. Extensive experiments across various models and benchmarks demonstrate that JECS achieves higher power than the max-p baseline while consistently maintaining the target GCR control.

cs.LG