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

+0新增论文
+0其中 AI 相关
+0已有论文修订
+91新增 AI 导读

arXiv 最近完成扫描:2026-11-22 14:51 · 全站最近入库:2026-11-21 12:19各信源进度 →

统计口径与数据新鲜度

统计 2026-11-21 16:31 — 2026-11-22 16:31(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:2026-11-22 14:51最近发现新增版本:2026-11-21 12:19库内论文最新更新时间:2026-11-20 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。

29,158 已收录论文24,181 AI 相关8,992 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 23,345 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-22 14:49(北京时间)
新论文 · v1arXiv:2611.25672

SEAF: Selective Semantic Enrichment and Visual Evidence Fusion for Multimodal Sarcasm Detection

Faxian Wan, Xiaocui Yang, Yifan Cao, Shi Feng, Daling Wang, Yifei Zhang

原文摘要

Multimodal sarcasm detection requires textual, visual, and contextual evidence, particularly when multiple images provide complementary cues. Multimodal large language models (MLLMs) can generate auxiliary information from a post's text and images, but their outputs may mix useful background knowledge with redundant visual descriptions. We propose SEAF, a dual-branch framework for selective semantic enrichment and visual evidence fusion. SEAF assigns generated fields to distinct roles: entity background enriches semantic input, while image descriptions supply alignment statistics. The semantic branch combines the original post text, background from a frozen MLLM, and optical character recognition (OCR) text. The aligned visual branch combines global and local visual features with statistics of cross-modal and inter-image similarity. Text-Anchored Smoothed Aggregation (TASA) weights images by their similarity to the original post text while preserving contributions from images with low literal similarity. Summing the branch classification scores integrates contextual semantics with visual evidence. Experiments on the multi-image MMSD3.0 benchmark show that SEAF outperforms all compared methods.

cs.CV · cs.CL

新论文 · v1arXiv:2611.25671

LATER: Learning Action-Conditioned Temporal Evidence Retrieval for Predictive World Models

Zihan Zhou, Wenxi Xu, Qilin Wang, Hao Tang

原文摘要

Early interactions can reveal physical properties that later observations no longer show. These properties can still gov- ern the effects of future actions. Inspired by the distinction between retention and selective readout in working mem- ory and cognitive science, we introduce LATER, an action- conditioned world model built on CF-JEPA. LATER stores completed real transitions in partitioned memories that remain fixed during candidate rollouts. A response bridge queries these memories using the current state and candidate action, while multi-cut auxiliary supervision trains the memory writ- ers. On distracted Two-Stage PushT, LATER raises terminal success from 15.5% to 25.0% over reconstructed CF-JEPA. In a separate comparison across six models and two scratch- training seeds, LATER achieves the lowest mean physical readout error in both clean and distracted views, reducing along-push mean absolute error (MAE) under distractors by 20.2% relative to CF-JEPA. Controlled history interventions show that later predictions depend on earlier physical evi- dence even when query observations and actions are fixed, with larger effects during pushing than waiting.

cs.AI

新论文 · v1arXiv:2611.25670

Efficient Latent Reasoning for Vision-Language Model via Dynamics-Aware Quantization and Distillation

Yanjing Li, Sheng Xu, Baochang Zhang, Xianbin Cao, Yawei Li

原文摘要

Efficient chain-of-thought (CoT) reasoning in vision-language models (VLMs) requires reducing both intermediate reasoning steps and their execution cost. Latent CoT methods compress textual rationales into continuous thoughts, yet their full-precision (FP) models retain substantial storage and computation costs. Lowering numerical precision could reduce these remaining costs, yet direct quantization severely degrades reasoning accuracy, particularly with 4-bit weights and activations. Operator and perturbation analyses reveal key sensitivity and direction-dependent predictive effects of continuous-thought errors, motivating preservation of information retrieval and subsequent reasoning beyond reconstruction accuracy. We therefore introduce QL-VLM, a dynamics-aware Quantization framework for efficient Latent reasoning in Vision-Language Models (VLMs). First, Dynamics-Guided Quantization (DGQ) uses nonuniform codebooks that preserve rotary-key structure, mitigating quantization-induced distortion in vision-language information retrieval. Second, Dynamics-Preserving Distillation (DPD) aligns continuous thoughts, latent-query readouts, and predictions with a full-precision teacher, maintaining predictive consistency through subsequent low-precision reasoning. On ScienceQA-IMG with Qwen2-VL-2B-Instruct, 4-bit QL-VLM achieves 90.43% accuracy and improves baseline by 40.33%, with 16.64x fewer bit-operations and 6.40x lower measured latency than explicit CoT.

cs.AI · cs.CV

版本更新 · v3arXiv:2603.24974

The Value of Information in Resource-Constrained Pricing

Ruicheng Ao, Jiashuo Jiang, David Simchi-Levi

原文摘要

Dynamic pricing with resource constraints requires balancing exploration of unknown demand patterns against exploitation of known information to maximize revenue. We develop algorithms for this problem across different information regimes. Products share perishable resources, and the firm posts prices and observes the realized requests, which may exceed the remaining capacity. In the full-information setting, a residual-correction policy that solves the fluid problem once attains $O(\log T)$ regret with bounded prices and arbitrary-rank resource matrices. In the no-information setting, deterministic exploration followed by a certainty-equivalent policy attains $O(\sqrt T)$ regret when the limiting fluid solution is nondegenerate. In the informed-price setting, given an initial price-demand estimate, a feedback controller learns the implementing prices without estimating the slope matrix if the slope matrix is symmetric, a set containing it is known in advance, the informed price is aligned with the resource constraints, and an inverse-image condition holds. A selector chooses between the informed-price algorithm and the no-information algorithm from historical data. Its regret bound includes observation error, misalignment, inventory exits, and periods on the fixed price. We further show how offline surrogate models reduce the regret of the learning policy through variance reduction, even when misspecified. The bound includes calibration and screen failures, and the regret is reduced for all sufficiently large horizons. We finally demonstrate the empirical performances of our policies through complete numerical experiments.

math.OC · cs.LG · stat.ML

版本更新 · v2arXiv:2611.20846

EDITORS Know Your Style! Editing LoRA Subspaces for Stylistic Attribution and Imitation

Zixuan Wang, Gregory Kang Ruey Lau, Xiao Tian, Jue Fan, Caroline Chaux, Bryan Kian Hsiang Low

原文摘要

Disentangling writing style from semantic content is a fundamental challenge in literary text modeling. Style-content entanglement causes models to rely on semantic content for authorship attribution (AA) tasks and memorize author-specific content for style imitation (SI) tasks. We propose EDITORS (EditingLoRA Subspaces), a diagnose-then-deflate framework that adapts pretrained large language models and prioritizes stylistic features, making it effective for both AA and SI tasks. EDITORS first trains a diagnostic LoRA adapter using style-neutral statements derived from the training corpus to identify a content-oriented subspace. It then trains the final style adapter under activation-space deflation that projects input activations away from the identified content-oriented directions. We empirically demonstrate that EDITORS is effective. On three AA benchmarks spanning literature, social media, and news, EDITORS achieves state-of-the-art accuracy, outperforming the strongest existing baseline by up to 29%. On SI, EDITORS generates stylistically faithful and diverse passages, outperforming baselines in achieving style similarity while adhering to content specifications.

cs.CL · cs.AI

新论文 · v1arXiv:2611.25667

CauSyncVLA: Causal Latent World Modeling for Asynchronous Robot Control

Xu Pan, Chubin Zhang, Xingrui Yu, Runmin Lan, Zihao Wang, Wenkai Guo, Bofang Jia, Bo An 等 9 位作者

原文摘要

Vision-Language-Action (VLA) policies must react while visual inference is still running. Overlapping computation and execution reduces waiting, yet leaves perceptual memory stale and successive action chunks disconnected. We present CauSyncVLA, an asynchronous dual-system controller that aligns perceptual-memory prediction and trajectory correction with actual robot execution. Rather than forecasting pixels or physical states, a Causal Latent World Model predicts current perceptual memory from delayed visual anchors and executed state-action history, supervised through action-sensitive distillation to retain the frozen expert's control response. Conditioned on this updated memory, Predictive Flow Warm-Starting (PFWS) aligns the unexecuted action horizon for single-step flow correction without extra training. On LIBERO-10, CauSyncVLA reaches 93.8% success at 5.04$\times$ inference throughput, improving closed-loop robustness and robot motion smoothness.

cs.RO · cs.CV

新论文 · v1arXiv:2611.25665

GNetSS: Towards Learning GNSS Corrections without Ground Truth

JiaXu Chee, Simon Denman, Fahira Afzal Maken, Qingqing Cheng, Amir Khodabandeh, Milad Ramezani

原文摘要

We present GNetSS, a self-supervised GNSS position correction framework that improves the robustness of single point positioning (SPP) without requiring ground-truth locations for training. GNetSS combines a transformer-based network with a deterministic differentiable positioning solver. The network predicts satellite-wise pseudorange corrections by modelling inter-satellite geometry and temporal context, while the solver converts these corrected measurements into receiver position corrections under GNSS geometry and optional motion constraints. Unlike supervised GNSS correction methods that rely on ground-truth, GNetSS derives its training signal directly from GNSS measurement consistency. Experimental results show that GNetSS improves robustness across open-sky, urban, and forested environments. The method is particularly effective at suppressing large-error outliers, reducing the maximum 3D positioning error by over 50% compared with SPP. These results demonstrate the potential of self-supervised pseudorange correction for scalable and robust GNSS positioning in real-world autonomy applications. Code and the in-house dataset will be released upon acceptance.

cs.RO

新论文 · v1arXiv:2611.25661

GRODE-RTL: Source-Mapped Dependency Views for LLM-Based RTL Repair

Jingyi Zhou, Yi Liu, Bin Sun, Hongji Zhang, Dongsheng Zuo, Qiang Xu

原文摘要

Large Language Models show great potential for automated Register-Transfer Level repair, yet they struggle with non-local faults where root causes and failing outputs reside far apart in source code. Treating RTL as a flat token sequence obscures hardware dataflows, whereas exposing complete synthesized graphs overwhelms the model with fine-grained metadata. To resolve this representation dilemma, we propose GRODE-RTL, an anchor-centric representation that extracts the single-clock-stage fan-in cone of a suspect register from the netlist, maps netlist nodes back to editable source lines, and organizes them into a compact hop-partitioned dependency view. By prioritizing direct drivers and folding distant hops, GRODE-RTL guides standard tool-using agents without exposing ground-truth hints or oracle feedback. We evaluate GRODE-RTL across 30 formally validated non-local faults using three LLMs (DeepSeek-V4-Flash, GLM-5.2, and MiniMax-M3) alongside a controlled six-configuration ablation study. GRODE-RTL improves pass@1 over source-level exploration by 16.7% on DeepSeek, by 26.7% on GLM-5.2, and by 20.0% on MiniMax-M3, with consistent gains on the hard-to-retrieve far-distance and indirect-anchor subsets of all three models. Our controls isolate structural relevance from generic context expansion, proving that selective dependency organization rather than full graph exposure is key to effective RTL repair.

cs.AR · cs.AI · cs.MA

版本更新 · v2arXiv:2611.22098

Every Measurement, Every Direction, All at Once: Multimodal Flow Matching for Molecules and Spectra

Thorben Prein, Elton Pan, Rafael Gómez-Bombarelli, Santiago Miret

原文摘要

Generative modeling in science is often data-limited, where experiments are expensive. However, each experiment may yield several coupled measurements of the same physical entity. Standard generative models often train only on a single direction, discarding supervision from the rest. We argue that each experiment contains more training signal than this asymmetry allows the model to exploit: random masking enables measurements to serve as either conditioning inputs or prediction targets, yielding multiple supervised denoising tasks from a single experiment. We investigate this premise with MOSAIC, a multimodal flow matching transformer over molecular 3D structure and UV, IR, and Raman spectra. MOSAIC is a single model that is capable of any-to-any generation: generating any subset of modalities from any disjoint subset of the others. On QM9S, MOSAIC achieves 88.8% molecular graph recovery, outperforming the strongest spectra-to-structure baseline by 23.2% in structure generation, while simultaneously outperforming a structure-to-spectra specialist model in spectral generation. Ablations show that the any-to-any objective improves structure generation over structure-only training across model sizes. Furthermore, MOSAIC supports arbitrary subsets of available measurements and enables molecular identification by matching generated spectra to experimental references.

cs.LG · physics.chem-ph

新论文 · v1arXiv:2611.25659

DNA: Unlocking Depth Scaling in CTR Models

Caiyuan Li, Wenjie Wu, Yuheng Huang, Xiao Zhou, Silu Zhou, Ke Chen, Zinan Zhao, Xinrui Zeng 等 17 位作者

原文摘要

Long user behavior sequences are among the strongest signals for click-through rate (CTR) prediction. Cross-attention makes these sequences practical for industrial ranking by keeping compute linear in sequence length. Yet the gains from stacking cross-attention layers taper off: on Taobao search, growing a cross-attention stack from 16 to 28 layers adds only 0.02 percentage points (pt) of AUC. We argue that the key-value side is the bottleneck. Each layer projects the same behavior sequence into keys and values, so adding depth gives the model more reads of a fixed sequence. We propose DNA (Dual-stream Nexus Attention), which lets the sequence itself evolve with depth. At each layer, the request representation (the search query, user features and scene) first reads the sequence. Each behavior token then attends to the updated request representation and uses what it reads to update itself, so the next layer reads a sequence shaped by the current request. The update runs once per request, all candidates share the resulting sequence, and both the reads and the update keep linear complexity in sequence length. At comparable forward FLOPs, DNA improves AUC over cross-attention stacks by 0.08 to 0.26 pt and keeps improving in the depth range where those stacks level off. DNA-4 is deployed in Taobao search. In an online A/B test against the production model, it improves CTR by 0.8% and GMV by 0.5%.

cs.IR

新论文 · v1arXiv:2611.25658

SiliconGym: A Versatile Reinforcement Learning Environment for Agentic EDA

Hejia Zhang, Sheng Lu, Yitong Ma, Zhongming Yu, Zhijing Wu, Wentao Ni, Chia-Tung Ho, Jishen Zhao

原文摘要

Reinforcement learning (RL) with electronic design automation (EDA) feedback is advancing large language models for hardware design and verification. Extending these gains to hardware-engineering agents requires environments that span tasks, preserve state across tool calls, and support large hierarchies alongside parallel rollouts. We present SiliconGym, the first versatile, cross-task and cross-toolchain RL environment for hardware design and verification. By decoupling versioned workspaces, budgeted execution, and acceptance-gated scoring, SiliconGym scales from module-level RTL, testbench, and PPA tasks to multi-round repository repair and full-hierarchy SoC flows. Streaming dispatch overlaps tool checks with generation, finishing a 256-answer RL batch 21.9% faster than native ChipSeek. Using interchangeable open-source and commercial backends, we show that open-source-trained functional correctness transfers cleanly to commercial simulation, while paired netlist evaluation reveals when open-source PPA gains survive commercial synthesis.

cs.AI · cs.LG

版本更新 · v3arXiv:2610.04425

AgroGround: What Grounding Fine-Tuning Changes in Agricultural Vision-Language Models

Abdulla Alshehhi, Zongyan Han, Rao Muhammad Anwer

原文摘要

Large-scale agricultural grounding data can now be generated automatically, but how training on such data affects a vision-language model's localization, recognition, and refusal capabilities remains unclear. To study these effects, we introduce AgroGround, containing 794,850 automatically grounded agricultural instructions for training and a separate benchmark of 1,478 human-verified images for evaluation. Among these images, 500 are independently re-annotated, yielding an inter-annotator instance F1 of 0.789. We use controlled comparisons to isolate the effects of different training choices on localization, recognition, their joint correctness, and refusal. Grounding-only fine-tuning reduces four-way recognition from 51.7% to 29.2%, with predictions showing a strong candidate-position bias. However, changing only the instruction format raises recognition to 72.7% while maintaining similar grounding performance. Training with healthy-image negatives achieves 94.1% refusal on healthy images, but only 28.6% refusal of incorrect target names on positive images, showing that these two refusal behaviors do not necessarily improve together. Reinforcement learning reveals a further trade-off, improving localization by 0.014 F1 over a matched supervised control while reducing recognition and refusal. Across model families and held-out training sources, localization transfers more reliably than recognition and target-specific refusal. External benchmarks further show that the gains depend on the domain and task. Together, these findings show that grounding fine-tuning changes the balance between localization, recognition, and refusal rather than improving all three uniformly.

cs.CV · cs.AI

新论文 · v1arXiv:2611.25654

DARSY: A Tool for Data-Driven Abstraction and Policy Synthesis for Stochastic Systems with Unknown Dynamics

Thom Badings, Ibon Gracia, Mahdi Nazeri, Morteza Lahijanian, Alessandro Abate

原文摘要

This paper presents DARSY: a software tool for abstraction-based verification and policy synthesis in discrete-time stochastic control systems with unknown dynamics and noise distributions. DARSY requires only a simulator of the system and a Lipschitz bound on its dynamics, and needs no knowledge of dynamics or noise distributions. The tool is fully automated and constructs a finite interval Markov decision process (IMDP) abstraction from sampled transitions, whose probability intervals are obtained via statistical techniques and Lipschitz reasoning. Robust dynamic programming on this IMDP then yields a policy together with a lower bound on the probability that the unknown system satisfies a desired specification. This lower bound holds with a user-specified confidence; that is, it is probably approximately correct (PAC). The tool is implemented in Python and supports general Lipschitz-continuous system dynamics and specifications expressed in syntactically co-safe LTL (scLTL), which includes reach-avoid and finite-horizon safety specifications as special cases. Beyond standard IMDPs, DARSY implements a two-layer IMDP (2IMDP) abstraction that also bounds the probability of transitioning to aggregated state blocks, yielding smaller abstraction errors than IMDP abstractions and, as a result, tighter satisfaction guarantees. We demonstrate DARSY's broad applicability across a variety of benchmarks and specifications.

eess.SY · cs.AI · cs.LO

新论文 · v1arXiv:2611.25653

ColorRef: Calibrating Verbal Feedback for Color Revision

Pablo Loyola, Andres Hoyos-Idrobo, Edison Marrese-Taylor, Colin Rowat, Alfredo Solano

原文摘要

People rarely specify an exact color when asking someone to revise a design: they might say \textit{a little darker} or \textit{more blue} after seeing a proposal. A language-model assistant must then infer not only which way to change the color, but how far. We study this distinction with \textsc{ColorRef}, a color reference game in which a model guesses a color from a description and a teacher, knowing a recorded target hidden from the guesser, supplies verbal corrections. On 1{,}000 crowd-sourced description--target pairs, three rounds of deterministic oracle feedback reduce mean target distance from 42.33 to 22.15 $ΔE_{76}$. However, those corrections provide additional information, and improvement alone does not establish well-calibrated feedback following. A controlled Qwen3-14B diagnostic reveals a more specific failure: following a requested direction and ordering graded phrases do not guarantee a suitable step size; strong modifiers can induce excessive changes. We therefore measure each receiver's typical response to \textit{a little}, \textit{somewhat}, and \textit{much} on separate colors and select wording matched to its observed steps. On 128 held-out starting colors, calibration reduces one-step mean error from 12.36 to 6.86 for Qwen3-14B and from 11.06 to 5.89 for Gemma 4 12B against a fixed graded-wording rule. The benefit is concentrated on large corrections, with many unchanged cases and no consistent increase in threshold success. Our findings show that following the direction of feedback is not the same as making an appropriately sized revision.

cs.CL

新论文 · v1arXiv:2611.25651

MatVerse: A Foundation Model for Materials Science

Xuanyu Zhu, Qian Tan, Bo Chen, Jiaqing Xie, Yang Han, Mei Wu, Wanhao Liu, Xuanze Lin 等 18 位作者

原文摘要

Materials science involves highly heterogeneous scientific information, including textual knowledge, molecular and crystal structures, experimental characterization data, microscopy images, and structure-property relationships. Developing a general-purpose foundation model capable of understanding and reasoning over these diverse materials representations remains challenging. In this technical report, we introduce MatVerse, a materials foundation model built upon Intern-S2-Preview and further specialized through large-scale materials-domain post-training. We construct a large-scale multimodal training corpus covering materials knowledge, scientific reasoning, quantitative problem solving, structural understanding, and multimodal scientific interpretation. To improve domain adaptation, we develop a unified data construction pipeline combining teacher-guided reasoning distillation, scientific problem synthesis, consistency-based filtering, and retrieval-based multimodal data construction. We evaluate MatVerse on three representative materials-science benchmarks: our human-calibrated OmniMatBench, MatSciBench, and MaCBench. Results demonstrate consistent improvements after materials-specific post-training over the base model on all three benchmarks, alongside improved inference efficiency on calculation-intensive problems. These results show that targeted domain-specific post-training can effectively enhance scientific reasoning, materials knowledge understanding, and multimodal scientific capabilities of foundation models. MatVerse provides a unified language-and-vision interface for materials science and serves as a foundation for future AI-assisted materials research workflows.

cs.AI · cond-mat.mtrl-sci

版本更新 · v3arXiv:2609.16726

Improved Separations between Quantum and Classical Communication Complexity of Total Functions

François Le Gall

原文摘要

We refine Gavinsky's framework (arXiv:2608.18784) for exponential separations between quantum and randomized communication complexity of total functions and obtain larger separations: polylogarithmic quantum communication versus $\tildeΩ(\sqrt n)$ randomized communication with two quantum messages, and versus $Ω(n^{1-\varepsilon})$ for every fixed $0<\varepsilon<1$ with more quantum messages. Our main technical contribution is a general conversion from quantum versus randomized communication separations for partial functions to separations for total functions. Our construction combines the communication lookup construction of Anshu, Belovs, Ben-David, Göös, Jain, Kothari, Lee, and Santha (FOCS 2016), the fully linear PCPs of Boneh, Boyle, Corrigan-Gibbs, Gilboa, and Ishai (CRYPTO 2019), and a linear seeded extractor with a short seed based on the construction of Raz, Reingold, and Vadhan (JCSS 2002). The key idea is to use the extractor to compress the lookup table into a single cell, thereby reducing the input length.

quant-ph · cs.CC

新论文 · v1arXiv:2611.25650

Extending Rademacher Complexity-Based Generalization Theory to Domain Generalization

Xi Wang, Liang Bai, Xian Yang, Richard Yi Da Xu

原文摘要

Domain generalization (DG) aims to train models on multiple source domains with distinct distributions to ensure robust performance on unseen target domains. In doing so, it elevates the concept of generalization from the traditional sample level to the distribution level. Existing theoretical frameworks introduce a domain space endowed with a meta-distribution to extend classical generalization analysis to this new setting. However, these approaches typically treat the domain space as devoid of geometry. In reality, the domain space is an inherently infinite-dimensional, nonlinear abstract space composed of probability distributions, and its complex geometric properties fundamentally determine generalization behavior. Consequently, existing geometry-free theories of domain generalization have a fundamental limitation. They lack a formal characterization of what distinguishes distribution-level generalization from sample-level generalization, leaving the inherent difficulty of generalizing across domains largely unexplained. To bridge this gap, we formalize the domain space as the Wasserstein space over the representation space. Building on the geometric properties of this space (specifically the $\mathrm{CD}(K,\infty)$ curvature theory) we derive a novel, geometry-dependent DG error bound. By introducing an additional curvature-dependent confidence term, this bound formally disentangles the two levels of generalization effects and reveals the essential nature that domain generalization is affected by both the invariance and the curvature of the representation space. Finally, we experimentally validate the effect of curvature on domain generalization.

cs.LG · cs.CV

新论文 · v1arXiv:2611.25648

TRACE: Attribute Execution for Sequential Image Editing

Xiyang Zhang, Yuanhe Tian, Hongzhi Wang, Yan Song

原文摘要

Given an image and successive user instructions, sequential image editing should return an updated image that satisfies each new request and preserves earlier requirements that remain active. This lets users develop a result without repeatedly restating or repairing their decisions. The difficulty is that a retrieved image can contain a valid color on an obsolete surface, while the current image may lack the newly requested material. Neither reference selection nor a precise object mask alone specifies how to reconcile attributes occupying the same region. We present TRACE, which separates the continuing attribute from its visual carrier and prioritizes local appearance operations under complete-request checks. Active-goal binding identifies what should persist, carrier selection supplies suitable current or newly generated pixels, and tonal calibration reduces tinting near highlights and shadows. Experiments across frozen editors, independent visual readers, new source images, and complete sequential benchmarks support joint color-material recovery and show how local execution preserves the surrounding scene.

cs.CV

新论文 · v1arXiv:2611.25646

AdsNet: A Graph Neural Network for Direct Adsorption Energy Prediction

Changzhi Ai, Frank Abild-Pedersen, Johannes Voss

原文摘要

Fast and accurate prediction of adsorption energies is critical for high-throughput computational catalyst screening, existing methods compromise either on evaluation speed (full-graph machine learning potentials) or geometric sensitivity (coordination-based models like the alpha-scheme). Here, we present AdsNet, an equivariant graph neural network that directly predicts adsorption energies in a single forward pass, which operates on a local graph restricted only to adsorbate atoms and its immediate surface neighbors within a cutoff radius. This local graph formulation decouples computational complexity from total system size while encoding continuous geometric features (interatomic distances and directional vectors) enabling sensitivity to subtle variations in local bonding environments. AdsNet adopts a message-passing architecture with scalar and vector node features over multiple interaction layers. Trained on AdsTraj, a dataset of 33,920 density functional theory (DFT) calculated structures spanning 26 transition metals and 6 adsorbates (H*, C*, O*, CO*, OH*, HOCO*), the model achieves an adsorption energy root-mean-square error (RMSE) of 0.032 eV and a force RMSE of 0.047 eV/A. Analytic forces obtained via automatic differentiation enable structure relaxation through a projected gradient descent algorithm with adaptive line search, eliminating entire geometry optimizations. Additionally, AdsNet uniquely decomposes predicted adsorption energies into contributions from individual atoms within the local adsorbate graph - an interpretability feature unattainable with conventional machine-learning potentials. AdsNet thus establishes an accurate, highly efficient, and interpretable framework for large-scale adsorption energy prediction and local interface relaxation across various material surfaces.

cond-mat.mtrl-sci · cs.LG · physics.comp-ph

新论文 · v1arXiv:2611.25644

MambaLIFT: Reference-Based Mechanism-Aware Attribution for VMamba

Jintu Kakati, Niladri B. Puhan

原文摘要

Attributing predictions of visual state-space models remains challenging because input-dependent selective dynamics couple feature selection, recurrent state propagation, and shared computational pathways. We introduce MambaLIFT, a reference-based, multiplier-first attribution framework for VMamba that propagates contributions through its computational graph using interaction-aware rules for selective-scan injection, recurrent memory, LayerNorm, and shared pathways. MambaLIFT decomposes the reference-relative class-logit difference, \(f_c(x)-f_c(x^0)\), by spatial position, VSS block, scan direction, and computational mechanism. Experiments on ImageNet-1K evaluate VMamba-T and VMamba-B, with hyperparameters selected on a calibration set and faithfulness assessed on a disjoint held-out evaluation set. MambaLIFT achieves numerical conservation to floating-point precision across audited attribution interfaces in both configurations. Under mean-colour pixel flipping, it achieves significantly higher mean faithfulness scores than MambaLRP on both configurations and the highest scores among the evaluated methods, including Integrated Gradients with matched references. Under blurred replacement, MambaLIFT is non-inferior to MambaLRP on VMamba-T and equivalent on VMamba-B, indicating that its advantage depends on the replacement protocol. Ablations on VMamba-T show that interaction allocation, temporal propagation, and pathway composition affect attribution maps. Mechanism-level analyses further show that exact relevance conservation does not imply additive effects under finite interventions: the mechanism decomposition accounts for the reference-relative logit difference but does not estimate intervention effects or establish causal validity.

cs.AI · cs.CV · eess.IV

新论文 · v1arXiv:2611.25643

TRAC: Efficient Diffusion Language Models for Machine Translation via Length-Aware Decoding

Tianyuan Li, Xihan Yang, Chenglong Wang, Yuzhang Wu, Jingbo Zhu, Tong Xiao

原文摘要

Machine translation with large language models requires both high translation quality and low generation latency. Accelerating translation while preserving quality is therefore essential for practical deployment. Diffusion large language models (dLLMs) can generate tokens at arbitrary target positions through iterative denoising, offering a new paradigm for parallel translation beyond sequential autoregressive decoding. However, existing methods still incur redundant computation: fixed target generation lengths waste computation on unused positions, while confidence-only commitment delays medium-confidence tokens whose predictions have largely stabilized. To address these issues, we propose TRAC, a training-free adaptive decoding method for diffusion machine translation. TRAC adapts the target generation length using source length and translation direction, with a conservative fallback to the maximum length for empty or unusually short outputs. It preserves high-confidence commitment while identifying temporally stable medium-confidence tokens for early commitment using cross-step top-1 persistence and approximate Jensen-Shannon divergence over Top-$K$ distributions. On bidirectional Chinese-English translation across FLORES-200, WMT23, and NTREX-128, TRAC reduces model forward calls by 24.9% to 36.8% and achieves sentence-level generation speedups of 1.91 to 2.19 times over Fast-dLLM, with only a 0.02-point decrease in macro-average COMET. These results demonstrate that adaptive generation lengths and stability-aware token commitment improve the efficiency of diffusion machine translation.

cs.CL

版本更新 · v2arXiv:2605.01427

SixthSense: Proprioception-Only Whole-Body Wrench Estimation for Humanoids

Xingzhou Chen, Xiayan Xu, Yan Ning, Jiyu Yu, Yizheng Zhang, Siyi Qian, Lingzhu Xiang, Jiahao Chen 等 11 位作者

原文摘要

Humanoid robots are entering the physical world at scale, yet are often criticized as oversized toys---good at dancing, but lacking the force-interaction capabilities for practical tasks. Reliable whole-body contact perception is difficult because floating-base dynamics make proprioceptive wrench recovery ambiguous, while model-based estimators require accurate dynamics and support-force inputs. We propose SixthSense, which uses conditional flow matching to model sparse spatiotemporal fields of link-wise contact gates and 6-D wrenches conditioned on proprioceptive and IMU histories. We train one estimator on pooled data from lying, standing, walking, and motion tracking. Across extensive simulation experiments, SixthSense outperforms learned and model-based baselines in overall contact estimation and remains robust to observation noise. We further establish a real-robot benchmark with our newly collected Unitree G1 dataset, where SixthSense achieves the strongest overall performance among learned estimators and requires approximately 50\,ms per inference. Controlled ambiguity experiments show that its samples recover multiple dynamically plausible contact modes with improved local physical consistency. We finally deploy SixthSense for collision detection and contact-triggered physical human--robot interaction.

cs.RO

新论文 · v1arXiv:2611.25640

Erasing the Wrong Thing: Conditional Attribute Leakage and Concept Erasure in Shared Foundation-Model Embeddings

Haijun He, Fei Li, Donghong Ji, Junjun He

原文摘要

Medical-image embeddings are increasingly shared in place of raw images for authorized downstream tasks, yet they may reveal sensitive attributes. Standard concept-erasure methods target population-level guardedness. However, we prove that, when binary task and sensitive labels are associated, global and task-conditional linear guardedness cannot both hold while a nonzero linear task signal is preserved, so global erasure must either eliminate that signal or retain task-conditional linear leakage. We further show how global erasure affects linear readers' discrimination between positive and negative samples with different sensitive attributes, deriving an approximate closed-form prediction of its utility cost from pre-erasure statistics. Motivated by this conflict, we introduce Task-Conditional LEACE (TC-LEACE), a closed-form variant of LEACE that enforces linear guardedness within each task stratum of its fitting data. We evaluate four vision foundation models on 85 CelebA attribute pairs spanning a broad range of associations. After LEACE, conditional leakage persists across all four encoders, the discrepancy between marginal and conditional audits increases with association strength, and predicted utility costs track observed losses. We then examine hospital chest radiographs using three foundation models, including two biomedical models, in which chest-tube status is strongly associated with diagnosis. Although a marginal attack on LEACE-erased embeddings underperforms a diagnosis-only reference, chest-tube status remains detectable within diagnosis strata. TC-LEACE recovers most of LEACE's linear task-utility loss while keeping task-conditional leakage within a prespecified margin of LEACE's.

cs.CR · cs.AI · cs.CV · cs.LG

新论文 · v1arXiv:2611.25639

LG-FiLM: Local-Global Mixture-of-Experts Modulation for RAW-to-RGB Reconstruction

Fei Li, Xilei Cao, Shichang Zhai, Kun Zhan

原文摘要

An image signal processor does more than reconstruct an image. It also has to decide how strongly to denoise, how to balance color, and which tone curve maps the scene onto the display, and these settings depend on the scene itself, the curve that suits a backlit sky is not the one that suits an indoor portrait. Most neural ISPs learn a single fixed transformation and apply it to every scene. We instead make the adaptation an explicit part of the network. Our model, LG FiLM, keeps one shared encoder decoder and places a small mixture of affine experts at its bottleneck; the experts scales, shifts, and mixing weights are predicted from local neighborhood evidence together with a global scene descriptor. Reconstruction stays in a single shared pathway and only the processing mode is switched. We evaluate LG FiLM on three datasets, Fujifilm UltraISP, Zurich RAW to RGB, and a COCO derived benchmark built with an inverse ISP. The model reaches the best reconstruction fidelity of the compared entries on the primary benchmark and transfers to the other two.

cs.CV

版本更新 · v2arXiv:2611.22317

SIDUMN: Self-Supervised Image Denoising Under Symmetric Multiplicative Noise

Nyan Htet Naing, Gyeongha Hwang

原文摘要

Self-supervised denoising is challenging when the corruption is multiplicative and signal-dependent. In this letter, we propose SIDUMN, a symmetric noisier framework for self-supervised denoising under mean-one symmetric multiplicative noise. Given noisy observations and the noise law, the method samples an independent synthetic noise factor from the same law and constructs two complementary noisier images by applying the factor and its reflection about one. The denoising network is trained to predict one noisier image from the other, and at inference the reconstructions of $K$ such antithetic pairs are averaged. We show that the population target of the training loss is the clean conditional mean minus a multiplicative cross-term. We further prove that, for any trained network, antithetic averaging never increases the expected squared error of a single noisier-input reconstruction. Numerical results validate the theoretical findings and the proposed method.

cs.CV

新论文 · v1arXiv:2611.25637

Beyond Easy Clients: Difficulty- and Improvement-Aware Aggregation for Federated Reinforcement Learning in Robot Navigation

Ye Lin Tun, Chu Myaet Thwal, Eui-Nam Huh, Choong Seon Hong

原文摘要

Federated reinforcement learning enables a fleet of robots to jointly learn a shared navigation policy without exchanging raw experience. In practice, however, the robots rarely operate in the same conditions: an open aisle is far easier to navigate than a cluttered room, and takes less learning to master. Standard federated aggregation rules ignore this. Uniform averaging weights every client equally, whether or not it is still learning, and weighting by episode return is worse, since easy environments permit high returns by construction. Both failures stem from what we call the \emph{plateau problem}: once a client masters its environment, its updates carry little new information even though its reported return remains high. We propose \emph{FedDIA}, which scores each client by how much its local training improves the global model it received, amplified by the difficulty of its environment. Difficulty is inferred from the model's own performance, without requiring any description of the environment. Both signals are measured against the global model, so they are comparable across clients. On a mapless navigation task with eight heterogeneous clients, FedDIA reaches a mean success rate of $75.5\%$ across training and held-out environments, against $68.7\%$ for the strongest federated baseline. Its advantage holds when the models are transferred to hand-authored warehouse and living-space layouts, or when one of the clients is defective.

cs.RO · cs.AI

新论文 · v1arXiv:2611.25636

Tenon: A Unified Programming Framework for Retargetable PIM Compilation

Niansong Zhang, Shiran Guo, Zhiru Zhang

原文摘要

Processing-in-memory (PIM) is emerging as a building block of heterogeneous systems for memory-bound workloads, and commercial prototypes now span near-bank SIMD engines, bank-level GDDR accelerators, programmable DRAM cores, and bit-level in-SRAM processors. Their programming stacks are fragmented: no abstraction or compiler is shared across vendors, and per-device models range from instruction sequences to C-like languages. Data layout further complicates the problem: tight compute-memory integration makes data layout a first-order optimization, yet existing PIM compilers lack a unified way to represent and optimize it. A key observation is that PIM compilation targets can be modeled as nested grids of compute and storage units connected by legal data-transfer paths. We present Tenon, a unified programming framework for retargetable PIM compilation. It decouples workload, compilation target, and cost specifications, so a workload is expressed once and compiled through a unified programming framework. Tenon generalizes the $\mathbb{F}_2$ linear layout algebra from GPU compilers to map logical work and data onto these compute and storage grids. It automatically constructs a minimum-movement data layout given a workload and a target specification. We evaluate Tenon on six PIM backends: five commercial backends and AttAcc, a state-of-the-art academic HBM-based PIM architecture. Across the commercial backends, it matches or outperforms vendor, expert, and compiler baselines, with geometric-mean speedups of up to $1.20\times$ and individual speedups up to $3.65\times$. For end-to-end GPT-13B decoding on AttAcc, Tenon-compiled workloads run $1.2$--$1.3\%$ faster than those compiled by DCC, a state-of-the-art PIM compiler, across all three batch-4 configurations. Together, these results demonstrate Tenon's performance portability across diverse PIM target families.

cs.PL

版本更新 · v4arXiv:2604.08014

Bridging Time and Space: Decoupled Spatio-Temporal Alignment for Video Grounding

Xuezhen Tu, Jingyu Wu, Fangyu Kang, Qingpeng Nong, Kaijin Zhang, Chaoyue Niu, Fan Wu

原文摘要

Spatio-Temporal Video Grounding requires jointly localizing target objects across both temporal and spatial dimensions based on natural language queries, posing fundamental challenges for existing Multimodal Large Language Models (MLLMs). We identify two core challenges: \textit{entangled spatio-temporal alignment}, arising from coupling two heterogeneous sub-tasks within the same autoregressive output space, and \textit{dual-domain visual token redundancy}, where target objects exhibit simultaneous temporal and spatial sparsity, rendering the overwhelming majority of visual tokens irrelevant to the grounding query. To address these, we propose \textbf{Bridge-STG}, an end-to-end framework that decouples temporal and spatial localization while maintaining semantic coherence. While decoupling is the natural solution to this entanglement, it risks creating a semantic gap between the temporal MLLM and the spatial decoder. Bridge-STG resolves this through two pivotal designs: the \textbf{Spatio-Temporal Semantic Bridging (STSB)} mechanism with Explicit Temporal Alignment (ETA) distills the MLLM's temporal reasoning context into enriched bridging queries as a robust semantic interface; and the \textbf{Query-Guided Spatial Localization (QGSL)} module leverages these queries to drive a purpose-built spatial decoder with multi-layer interactive queries and positive/negative frame sampling, jointly eliminating dual-domain visual token redundancy. Extensive experiments across multiple benchmarks demonstrate that Bridge-STG achieves state-of-the-art performance among MLLM-based methods. Bridge-STG improves average m\_vIoU from $26.4$ to $34.3$ on VidSTG and demonstrates strong cross-task transfer across various fine-grained video understanding tasks under a unified multi-task training regime.

cs.CV

新论文 · v1arXiv:2611.25632

What One Sample Cannot See: Why Fewer Denoising Steps Can Score Higher on Physics Benchmarks

Yutao Li, Chen Wang

原文摘要

When a storm can head north or south, a forecast down the middle may score better under squared error than either plausible path. The same paradox appears when video generators are tested as world models: fewer denoising steps can raise physics scores even as videos blur or stop moving early. Why? Physics-IQ compares each generated video with one recording, although a scene can have multiple valid futures. An exact decomposition shows its scores can reward reduced variation among generated futures, even when their distribution becomes less accurate. We introduce Kernel Physics-IQ to remove this incentive in expectation, needing only two independent samples per condition. Two-step sampling lowers Kernel Physics-IQ in all five image-to-video models. For CogVideoX, the official score rises by 0.055 while ours falls by 0.131. Changing the sampling schedule reveals further differences official scores miss. In controlled physical systems with known outcome probabilities, our metric favors the true distribution over collapsed predictions. In short, a world model should be judged not only by how closely one video matches what happened, but by how faithfully its predictions represent what could have happened.

cs.CV · cs.LG

版本更新 · v2arXiv:2611.02011

MedPsy: State-of-the-Art Small Medical Language Models for Efficient Edge Deployment

Davide Vitabile, Alexandro Buffa, Akshay Nambiar, Amril Nazir

原文摘要

Medical language models promise expert clinical support, but the strongest open-source LLMs remain too large for private edge-device deployment, while compact medical models still trail substantially on knowledge-intensive and real-world healthcare tasks. Reasoning models that fit edge memory budgets can produce long reasoning traces, increasing sequential decoding, end-to-end latency, energy consumption, and KV-cache memory pressure. We present MedPsy, a family of text-only 1.7B and 4B medical language models designed for edge deployment, targeting this accuracy-generation-efficiency trade-off. Our recipe combines a synthetic medical data pipeline over biology, medicine, and health seeds with chain-of-thought targets from a 235B medical-focused LLM teacher; a four-stage post-training curriculum comprising two SFT stages followed by two RL stages with hard-sample mining; and an edge deployment evaluation spanning multiple GGUF quantization variants and real-device inference on laptop and smartphone hardware. On seven closed-ended medical benchmarks, MedPsy-4B surpasses MedGemma-1.5-4B by 19.34 points and matches MedGemma-27B while being 6.75x smaller; MedPsy-1.7B outperforms MedGemma-1.5-4B by 11.42 points despite being less than half its size. On HealthBench-Hard, MedPsy-4B surpasses MedGemma-27B by 15.33 points, while MedPsy-1.7B surpasses it by 11.66 points despite being ~16x smaller. Beyond accuracy, MedPsy reduces average response length by 1.7x (1.7B) and 3.2x (4B) versus its Qwen3 backbones, while Q4_K_M quantization preserves the aggregate score within 1 point of BF16 with ~69% lower disk footprint. In real-device inference, MedPsy-4B's shorter generations translate into 75% lower median end-to-end latency and estimated energy per answer than its Qwen3 backbone at matched quantization, while peak RAM remains below 5 GB across our tested laptop and smartphone configurations.

cs.CL