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

+479新增论文
+361其中 AI 相关
+63已有论文修订
+93新增 AI 导读

arXiv 最近完成扫描:2026-11-16 01:34 · 全站最近入库:2026-11-16 13:44各信源进度 →

统计口径与数据新鲜度

统计 2026-11-15 15:58 — 2026-11-16 15:58(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:2026-11-16 01:34最近发现新增版本:2026-11-16 13:44库内论文最新更新时间:2026-11-14 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。

23,827 已收录论文19,941 AI 相关6,921 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 23,827 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-16 01:33(北京时间)
新论文 · v1arXiv:2611.13701

Topological Dynamics Under Non-Stationary Supervision: Persistent Homology of Neural Network Weight Evolution

Mitra Varun

原文摘要

We investigate how the persistent topology of neuron- weight point clouds evolves under stable and non-stationary supervision. An identical multilayer perceptron is trained on Fashion-MNIST with correct labels, a fixed random permutation of the training-label vector, or a fresh permutation every epoch. Five paired seeds share initialisation and minibatch order, yielding 15 models trained for 100 epochs. Raw and row-normalised incoming weight clouds are analysed with H0 and H1 Vietoris Rips persistence at 22 checkpoints. Correct-label models reach 89.39% mean true-label test accuracy; fixed-random models fit 67.94% of their assigned training targets without generalising to the original task. Correct-label models have descriptively greater time-averaged normalised fc3 H1 persistence than either random-label condition, but the configured exact comparisons do not reach conventional significance after Holm correction. Early condition classification achieves 60.0% mean balanced accuracy using topology, 66.7% using conventional diagnostics, and 60.0% using both. Thus, no incremental predictive improvement is demonstrated in this setting. The archive contains all 3,300 configured topology cases. Interpretation is limited by five seeds, a single architecture and dataset, 282 capped H1 distance comparisons, and deferred bottleneck analysis.

cs.AI

新论文 · v1arXiv:2611.13700

CJRR-Net: A Candidate-Joint Reprojection Ranking for Future-View Selection in Human Pose Estimation

Huiting Xi, Marsha Mariya Kappan, Eduardo Benitez Sandoval, Francisco Cruz

原文摘要

Multi-camera systems support performance recording, sports and human-movement analysis, where selecting an informative viewpoint is important for reliable human pose estimation. We propose the Candidate-Joint Reprojection Ranking Network (CJRR-Net) for future-view selection among eight calibrated cameras in MPI-INF-3DHP. The model combines current-view pose observations, camera configuration, and candidate-specific geometric features obtained by reprojecting the estimated 3D pose and its uncertainty into each candidate view. CJRR-Net achieves a selected-view 2D mean per-joint position error (MPJPE) of 105.18 pixels, compared with 114.64 pixels for the shared-input MLP and 117.56 pixels for adapted PoseErrNet. It also achieves the lowest mean error among the listed policies in each evaluated occlusion group. A blender rollout illustrates repeated camera selection. These findings support candidate-specific geometry for single-view selection in the evaluated fixed eight-camera setting.

cs.CV

版本更新 · v3arXiv:2504.18598

BadMoE: Backdooring Mixture-of-Experts LLMs via Routing-Aware Triggers and Task-Underutilized Experts

Qingyue Wang, Qi Pang, Xixun Lin, Shuai Wang, Daoyuan Wu

原文摘要

Mixture-of-Experts (MoE) architectures are rapidly becoming the standard for building scalable and efficient large language models (LLMs), thanks to their routing mechanism that activates only a subset of experts for each input. However, the modular structure and sparse activation of experts also introduce new security risks, as manipulating a small portion of experts may influence the behavior. In this work, we study a white-box, task-specific model-supply-chain setting and propose BadMoE, a routing-aware backdoor attack against MoE LLMs. We motivate BadMoE through a routing-aware decomposition of the MoE-layer output, which separates the contributions of the selected target experts from those of the remaining experts. This analysis highlights two measurable factors underlying the attack: the routing mass assigned to the target experts and their relative contribution to the layer output. Guided by this insight, BadMoE selects task-underutilized experts and utilizes routing-aware triggers to activate them, enabling reliable activation of task-specific, predefined backdoor behaviors. Specifically, BadMoE involves three steps: 1) identifying experts with low usage on a target-task calibration set, 2) optimizing a routing-aware trigger toward these experts, and 3) associating the selected-expert activations with attacker-specified behavior. We evaluate BadMoE on Mixtral, OLMoE, and Deepseek across sentiment, topic, and emotion classification, sentiment steering, and targeted refusal.

cs.CR · cs.AI

新论文 · v1arXiv:2611.14589

CAFIN: CUDA-accelerated Adaptive Fisher Information Matrix Network Optimization via GPU-Native Kernel Fusion

Maryam Arshad, Muhammad Khuram Shahzad

原文摘要

Second-order optimizers improve deep neural network training using curvature information, but their practical efficiency is limited by GPU execution overhead. AdaFisher reduces Fisher Information Matrix (FIM) approximation costs through diagonal Kronecker structures and exponential moving averages, yet its PyTorch implementation requires sequential CUDA kernel launches, intermediate tensor allocations, and repeated memory accesses. We introduce CAFIN, a GPU-native CUDA kernel fusion framework that improves AdaFisher execution efficiency without changing its optimization formulation. The proposed approach progressively integrates FIM preconditioner construction, exponential moving average updates, gradient preconditioning, momentum accumulation, weight decay, and parameter updates into fused CUDA kernels. Shared-memory reuse, register-resident computation, and in-place operations further reduce memory traffic and intermediate allocations. Experiments using ResNet-18 on CIFAR-10, CIFAR-100, and Tiny ImageNet across NVIDIA GTX 1660 Ti and RTX 3060 Laptop GPUs demonstrate up to 78.4% lower isolated FIM construction latency and 2.1-26.9% lower optimizer-step latency. The implementation also eliminates approximately 18.91 MB of intermediate GPU memory allocations per optimizer step. Numerical validation confirms consistency within the predefined FP32 tolerance, while convergence accuracy remains comparable across evaluated datasets and random seeds. These findings demonstrate that CUDA kernel fusion can substantially accelerate second-order optimization while preserving AdaFisher's underlying mathematical formulation and learning behavior.

cs.LG · cs.AI

版本更新 · v3arXiv:2611.05726

Do Language Models Know Which Constraints Belong Together?

Junbo Jacob Lian, Kaixuan Chen, Hanzhang Qin, Huiling Chen, Chaoyu Zhang, Chung-Piaw Teo

原文摘要

Dantzig--Wolfe reformulation strengthens the linear relaxation of a mixed-integer program by convexifying constraint families, and each candidate costs a column generation run. The choice is hard when families strengthen the bound only together, as in classical lot-sizing, network-design, and facility-location formulations. We prove that when an evaluation convexifies at most $s$ of $k$ families, methods that observe only bounds need of order $k^w$ evaluations to find a hidden group of $w$ families in the worst case. A prediction missing $e$ members reduces this to $Θ(k^e)$ through a staged search that need not know $e$. The orders extend to several groups, prices, and growing evaluation costs. We study two predictors: a lookahead at an optimal LP dual, which vanishes on groups whose families share no variable, and language models reading the model, verified by the staged search in JointDW. Such groups appear in three of ten MIPLIB-NL models whose bound can be improved and in all 59 benchmark instances. Given the coefficients of models with up to 16 families, one language model names a true group first in 249 of 252 answers. On the benchmark instances, decompositions chosen by language models from the text solve faster than those of GCG's detector.

cs.LG · cs.AI · cs.DS · math.OC

版本更新 · v2arXiv:2607.21067

PrefReward: Learning User Preference Matrix for Personalized Text Generation

Yue Wu, Chengbing Wang, Yimeng Bai, Xiaoyan Zhao, Yang Zhang, Fuli Feng

原文摘要

Large Language Models (LLMs) have demonstrated remarkable ability in generating personalized content by leveraging user histories and contextual cues. However, most existing personalization approaches rely on implicit representations within model parameters, making it difficult to interpret user-specific preferences or effectively handle long-context dependencies. To address these challenges, we propose PrefReward, a novel preference-aware generative framework that explicitly models user styles through a structured preference matrix and integrates it into the decoding process as a reward signal. PrefReward consists of two stages: (1) extracting a user-specific preference matrix that summarizes individual stylistic tendencies, and (2) using the matrix to guide generation via a KL-divergence-based reward function. Experiments on the LongLaMP dataset show that PrefReward outperforms retrieval-based baselines and context steering method in both generation quality and personalization interpretability.

cs.CL

新论文 · v1arXiv:2611.13697

Gödel's Incompleteness Theorems in Grounded Arithmetic

Bryan Ford

原文摘要

Reflective Grounded Arithmetic (RGA) is a paracomplete arithmetic in which every term is a program and truth is grounded evaluation. RGA discharges a hypothesis only when it proves the hypothesis decided. This paper shows, in a machine-checked Isabelle/HOL development, how Gödel's incompleteness theorems hold for RGA and for systems built on it. RGA's implication is material and RGA has the disjunction property, so the implications the textbook proof of the second theorem uses fail while its entailments and rules hold: Gödel's proof survives as entailments. RGA does not prove its consistency sentence Con, and RGA derives its own second theorem as the entailment $\mathit{Con}\vdash\neg\mathit{Pr}(\ulcorner\mathit{Con}\urcorner)$ and as one sentence quantified over proofs. The second theorem is then proved once, for every Gödel extension of RGA: an entailment relation on RGA's language, closed under RGA's rules, with a checker that meets three structural conditions, NEC, D3 and EXPL as proof transformers, D1 as closed acceptance, and consistency. D2 is not among the conditions. RGA with its checker is one instance. Inductive Grounded Arithmetic (IGA), RGA with ground induction, is another, its checker adding one rule detector, and IGA does not derive its consistency sentence for it. Adoption, a reflection rule whose premise is a derived warrant that the checker accepts computed proofs of every instance of a universal, gives consistent extensions RGA$^+$ and IGA$^+$. RGA$^+$ derives RGA's consistency sentence and IGA$^+$ derives RGA's and IGA's, so each strictly extends its base, and IGA$^+$ both. The second theorem holds at every adopted notation claim, and adopting Turing's code for that claim's checker derives its consistency sentence, so no claim certifies itself. Whether IGA alone derives RGA's consistency sentence is open, and the paper names what two routes to an answer need.

math.LO · cs.LO · cs.PL

新论文 · v1arXiv:2611.13696

On Private Knowledge Retrieval and Private Inference for secure multi-agent systems

Vicenç Torra, Maria Bras-Amorós

原文摘要

Multi-agent systems provide a convenient framework for the massive deployment of agents based on Large Language Models (LLMs). It provides a sound framework where agents can interact with existing software to perform actions in an autonomous way. Agentic AI is the new term to denote these AI-based autonomous systems. In real multi-agent systems, agents need to deal with malicious agents some of which can exploit security vulnerabilities to do harm. A memory poisoning attack is an example of exploiting a vulnerability to modify agents knowledge and, thus, affect its behavior. In an agent, there are different types of memory systems (semantic, episodic, and short-term) which have their own characteristics and for which different memory poisoning attacks may be developed. In this paper we focus on semantic memory, we discuss its attack, and mitigation strategies. We introduce private knowledge retrieval and private inference. We propose a signature system in the knowledge base to prevent from undesired manipulation. This should avoid memory poisoning attacks by introducing security by design in the construction of the semantic memory. In our architecture, an agent contains a local Prolog-like inference system (i.e., using Horn clauses to represent knowledge) which accesses an external knowledge base to get the required knowledge. This externally signed knowledge base is accessed privately. We provide different implementations in Python using different privacy levels including k-anonymity and information-theoretical privacy (IT-PIR).

cs.CR · cs.AI

新论文 · v1arXiv:2611.13694

Decision-Focused Learning via Neural Optimization Layers for Battery-Assisted Grid-Import Smoothing in AI Data Centers

Xinda Zheng, Canchen Jiang, Frits de Nijs, Edward Lam, Markus Wagner

原文摘要

Rapid AI data-center power fluctuations challenge grid integration. Sustained grid-import smoothing requires allocating limited battery power and energy under uncertain demand. We formulate risk-aware BESS scheduling that coordinates advance reserve commitments and demand-responsive dispatch, accounting for operating costs, grid-ramp penalties, and their conditional value-at-risk. We develop battery scheduling decision-focused learning (BS-DFL) with a constraint-aware neural optimization layer trained directly from this objective without optimal-solution labels. The trained layer is then frozen to guide demand-predictor learning through scheduling costs and ramping risk, enabling solver-free rolling-horizon scheduling. In a 10,000-minute simulation using loads derived from measured AI workloads, BS-DFL reduces mean absolute minute-to-minute grid-import changes by 71.70% and total scheduling cost by 18.01% relative to an MSE-trained predictor coupled with a mixed-integer linear programming solver, with an average decision time of 18.77 ms.

eess.SY · cs.LG · math.OC

新论文 · v1arXiv:2611.13693

Rotating Walks, Derandomized Squares, and Permutation Conjugated Walks

Vedat Levi Alev, Shravas Rao

原文摘要

Using the recent strong convergence results of Bordenave and Collins (Annals of Mathematics) and Chen et al. (Annals of Mathematics), we study the average case behavior of rotating random walks of Cohen and Maor (STOC 2023) and the derandomized square construction of Rozenman and Vadhan (RANDOM 2005). Concretely, we prove a conjecture of Cohen et al. (ITCS 2025) regarding the average case behavior of derandomized squaring for random graphs and show that the existential bounds of Cohen and Maor for their rotating random walk construction hold with high-probability for random graphs. Using the results of O'Donnell and Wu (FOCS 2020), we can in polynomial time construct a family of graphs and rotations which almost saturate the bounds of Cohen and Maor, and meet the bounds conjectured by Cohen et al. for all vertex-transitive auxiliary graphs of fixed degree in the derandomized square. Finally, we study independent random permutation conjugates of deterministic regular graph adjacency matrices. We prove quantitative weak convergence to freely independent tree adjacency operators when the degrees are fixed and the girths tend to infinity. Separately, a covering argument gives deterministic spectral lower bounds for polynomials with non-negative coefficients, recovering the lower bound of Cohen et al. for rotating walks.

math.CO · cs.DM · math.PR

新论文 · v1arXiv:2611.13691

Seeing Without Seeing: A Methodological Review of Visual Mental Imagery Reconstruction from Brain Activity

Shiyi Zhang, Dong Liang, Yihang Zhou

原文摘要

Reconstructing visual mental imagery from brain activity differs from reconstructing perceived images. Imagery is generated top-down, its signal is weak, and the imagined content is never observed, so decoders are trained and scored against a proxy such as a cue or a seen exemplar. This criterion-guided methodological review of mainly non-invasive studies (fMRI, EEG, MEG) is narrative rather than systematic. We organise the field into a two-stage pipeline, two design axes and four method families, and formalise target-free evaluation, in which the proxy gap is a systematic bias rather than noise. A two-dimensional admissibility criterion crosses trial-level separability from working-memory maintenance with the memory source of the imagined content. Among the studies reviewed, every image-scored reconstruction concerns a recalled exemplar, none builds a scorable target for concept-generated content, and all come from fMRI, consistent with missing image-level targets in electrophysiological data. Published values for five decoders, compared under a common human forced-choice framing, show seen-to-imagined drops of 5.9 to 26.1 percentage points and a change in method ordering. We treat this as a diagnostic signal, not a controlled benchmark result, and specify a within-subject test for an altered neural code. Beyond this synthesis, we make two proposals. First, because frozen generative priors can produce photorealistic images weakly tied to neural signals, we combine existing null-signal controls into a normalised neural dependence with a within-category null, an unvalidated diagnostic quantity. Second, we propose a candidate ten-item reporting checklist with illustrative pass criteria. In an exploratory application to ten studies, none reports all items in a fully checkable form; this describes reporting practice, not study quality.

cs.HC

版本更新 · v4arXiv:2604.08567

Multi-User Large Language Model Agents

Shu Yang, Shenzhe Zhu, Hao Zhu, José Ramón Enríquez, Di Wang, Alex Pentland, Michiel A. Bakker, Jiaxin Pei

原文摘要

Large language models (LLMs) and LLM-based agents are increasingly deployed as assistants in planning and decision making, yet most existing systems are implicitly optimized for a single-principal interaction paradigm, in which the model is designed to satisfy the objectives of one dominant user whose instructions are treated as the sole source of authority and utility. However, as they are integrated into team workflows and organizational tools, they are increasingly required to serve multiple users simultaneously, each with distinct roles, preferences, and authority levels, leading to multi-user, multi-principal settings with unavoidable conflicts, information asymmetry, and privacy constraints. In this work, we present the first systematic study of multi-user LLM agents. We begin by formalizing multi-user interaction with LLM agents as a multi-principal decision problem, where a single agent must account for multiple users with potentially conflicting interests and associated challenges. We then introduce a unified multi-user interaction protocol and design three targeted stress-testing scenarios to evaluate current LLMs' capabilities in instruction following, privacy preservation, and coordination. Our results reveal systematic gaps: frontier LLMs frequently fail to maintain stable prioritization under conflicting user objectives, exhibit increasing privacy violations over multi-turn interactions, and suffer from efficiency bottlenecks when coordination requires iterative information gathering.

cs.CL · cs.MA

新论文 · v1arXiv:2611.13690

Joint-State Inference for Autonomous Excavators Using Cab-View Vision-Control Fusion

Junwei Li, Jian Pan, Mengchi Cai, Shuai Zhao, Xizhuo Jiang, Zheng Wang, Qing Xu, Keqiang Li 等 9 位作者

原文摘要

Reliable state feedback of excavator front-end mechanisms is essential for autonomous excavation, yet retrofitting existing machines with dedicated joint sensors, external cameras, or visual markers can increase deployment and maintenance burdens. This paper proposes a closed-loop autonomous excavation framework for continuous inference of excavator joint states using onboard visual observations and control-related temporal information. The framework encodes structural cues from the current cab-view image, models short-term motion trends from historical joystick commands, and introduces a rolling state buffer for the bucket joint to preserve temporal continuity under severe occlusion. Mechanical joint-range constraints are incorporated into the learning objective to suppress infeasible estimates. The current cab-view image and inferred states are supplied to an Action Chunking with Transformers (ACT)-based control policy for online action generation. Experiments in an AGX Dynamics for Unity simulation environment under diverse operating conditions evaluate joint-state inference and end-to-end task execution. Across 20 autonomous excavation trials, the framework achieved a 95% task-completion rate, with an average joint-state-inference-and-control latency of 10.11 ms per step. The bucket-joint inference achieved a mean absolute error (MAE) of 0.965 degrees. These results demonstrate a sensor-light and computationally efficient joint-state-inference approach that can support closed-loop autonomous excavation with reduced dependence on additional sensing infrastructure.

cs.RO

新论文 · v1arXiv:2611.13689

Walrus Technical Report

Alkis Koudounas, Osamu Take, Hayato Futami, Quentin Jodelet, Emiru Tsunoo

原文摘要

We present Walrus, a unified 8B speech-and-audio foundation model handling speech recognition (including code-switching and multi-talker speech), speech-to-text and speech-to-speech translation, text-to-speech, emotion recognition, audio captioning, and general audio understanding. Our main contribution is not a new architecture but a data-centric training recipe; we use a standard decoder-only backbone initialized from Gemma 4, so that gains can be traced to the data. For speech generation, Walrus predicts low-bitrate discrete speech tokens, and a lightweight token-to-waveform decoder converts them into high-fidelity audio. The recipe is a three-stage data curriculum that moves from broad multilingual data to multi-talker, code-switching, and domain-specific data. At each stage, task-aware and dataset-aware weighting balances tasks with very different token budgets and protects core skills from forgetting. Trained on over 600M samples covering more than 20 tasks and 100 languages, Walrus leads on most multilingual, multi-talker, and code-switching ASR, speech translation, and audio understanding benchmarks, outperforming both open speech language models of comparable size and dedicated task-specific models. On most benchmarks, it even matches or surpasses the $4\times$ larger Qwen3-Omni's performance. Results suggest that data engineering is a primary driver of quality in multi-task speech and audio modeling.

eess.AS · cs.CL

新论文 · v1arXiv:2611.13686

You Are What You Read: Indian Political News Media Bias Dataset for Hindi Language

Kushal Trivedi, Mayank Mahor, Dhruv Sanghvi, Roshni Chakraborty

原文摘要

Media bias understanding and detection remains an under-researched problem for Indian context, specifically for Indic languages, such as, Hindi, which is widely spoken in India. In this paper, we introduce a comprehensive, multi-annotator, explainable dataset consisting of 448 unique articles from four leading Indian news agencies publishing in Hindi. We focus on Hindi, the most spoken and an official language of India. The study of media bias is particularly relevant in the Indian context due to the country's linguistic diversity and large population, where the same event is often reported by multiple news agencies in different languages, thereby introducing scope for bias. The dataset comprises Hindi political news articles annotated by 14 experts and classified into four bias categories - selection bias, framing bias, clickbait bias and omission bias - when an article is identified as biased or unbiased. For each biased article, annotators also provide a textual explanation, enabling future semi-automated approaches to incorporate explainability when language models reason over the dataset. Cohen's Kappa scores were 0.81 for simple bias/no-bias classification reflecting strong inter-annotator agreement. For the multi-label task, the Cohen's Kappa values are 0.89, 0.75, 0.81, and 0.79 for clickbait, selection, framing, and omission biases, respectively, resulting in an average Kappa of 0.80, which also reflects strong agreement.

cs.CL

新论文 · v1arXiv:2611.13687

AURA: Auditory Spatial Intelligence via a Unified Representation from a General-Audio Encoder through Complete Alignment

Boyan Han, Xiaoyan Wang, Zhe Sun, Ziqi Wang, Yujun Cai, Hongli Cao, Yiwei Wang, Ming-Hsuan Yang 等 9 位作者

原文摘要

Answering spatial questions about multi-source scenes requires a model both to identify each sound and to locate it. Existing spatial audio-language models obtain spatial information from a dedicated encoder, which separates spatial and semantic processing into distinct feature spaces. We study whether a general-audio codec trained without spatial supervision can serve both purposes. A controlled layer-wise analysis shows that the frozen codec retains spatial information, with lateral cues carried by the difference between the two ear channels. This information, however, is not readily accessible: in dual-source scenes, a model trained on these features predicts left/right at chance and, on the remaining axes, assigns the detected direction to the queried and the non-queried source equally often. We attribute this failure to an audio--text information asymmetry, in which the audio contains both sources while the answer describes only the queried one. A controlled $2\times2$ experiment shows that either completing the scene description or exposing the inter-channel difference lifts left/right accuracy above chance, whereas binding directions to the queried source requires both. Building on these findings, AURA keeps the codec and the language model frozen and exceeds the strongest dedicated system on SpatialSoundQA by 10.31 points in dual-source direction accuracy, with about 40\% of its trainable parameters.

cs.SD · cs.AI

新论文 · v1arXiv:2611.13685

Ego Guidance: Enhancing Robot Policy Learning with Action-Free Human Egocentric Videos

Lin Long, Tianren Ma, Zhong Chu, Di Wu, Maosong Cao, Qixiang Ye, Junbo Zhao, Shanghang Zhang

原文摘要

In this paper, we investigate how human egocentric videos can benefit robot policy learning without action supervision. Building on a pretrained VLM, we propose Shared Prediction on Action and Representation (SPAR), an MoT architecture in which future visual state prediction can provide independent and complementary learning signals alongside action prediction. We find that representations learned from egocentric videos improve downstream performance on out-of-distribution robot tasks, with further gains as video data scales. We also study the effects of objective choices and data curation, providing insights into how to leverage action-free videos for scalable robot learning.

cs.RO · cs.AI · cs.CV

版本更新 · v2arXiv:2605.28063

Unified Synthesis of Compositional Speech and Sound from Free-Form Text Prompts

Yuyue Wang, Xihua Wang, Xin Cheng, Yijing Chen, Ruihua Song

原文摘要

Generating unified audio that naturally integrates speech and sounds from free-form text prompts remains challenging. For example, the prompt ``A fast drum roll ends with a cymbal crash, after which a man shouts: `Come on!'.'' describes several events that should appear in order and form a coherent recording rather than independent tracks that lack coherent temporal and semantic interactions. Current methods broadly follow two paradigms: pipeline methods couple separate text-to-sound and text-to-speech models, while unified models rely on a single generator that still requires structured, multi-field prompt inputs. Both paradigms require upstream prompt decomposition or rewriting, limiting free-form flexibility and risking cascading errors. To enable direct generation from free-form prompts within a single model, we introduce PlanAudio, an end-to-end framework that maps unconstrained free-form text to unified audio without intermediate prompt preprocessing. To address the challenges of direct free-form generation, PlanAudio integrates latent planning into a single autoregressive framework, predicting continuous audio layout before acoustic token synthesis. We construct a large-scale training corpus and introduce Free-Form Unified Audio Generation Benchmark (FUAG) for evaluation, both spanning speech, sound, and joint speech-and-sound generation. Experiments show that a single PlanAudio outperforms existing pipeline and unified baselines on speech-and-sound generation, while remaining competitive with specialist models on speech and sound generation. Further analysis validates the effectiveness of audio layout planning over textual planning, acoustic planning, and direct generation. Sample Available: https://planaudio.github.io/PlanAudio2026/.

cs.SD · cs.AI · cs.MM

新论文 · v1arXiv:2611.13684

BypassGym: A Benchmark for Evaluating AI Agents' Endpoint Defense Evasion Capabilities

Weifeng Yuan, Zifan Chen, Yucheng Zhang, Feng Dong, Peng Gao, Haoyu Wang

原文摘要

Artificial intelligence agents increasingly automate cybersecurity tasks, including offensive operations. Frequent attacks across the web make endpoint detection and response (EDR) a critical defense, but whether AI agents can evade it remains unclear. Benchmarks evaluate the offensive capabilities of large language model (LLM) agents in areas such as penetration testing and vulnerability discovery. One important capability that these benchmarks do not address is defense evasion, specifically whether an agent can develop a program that executes a payload without being blocked or terminated by a deployed EDR. We introduce BypassGym, to our knowledge the first benchmark to systematically evaluate this capability on real endpoints. Given command-and-control (C2) shellcode, an agent must generate and compile a loader that executes the payload without being blocked or terminated. A supplementary evasion knowledge skill provides domain knowledge and supports comparisons with and without this input. BypassGym is built on Proxmox VE(PVE), so the evaluation environment runs real Windows systems and EDR products inside virtual machines and returns structured detection feedback to the agent. We evaluate 5 lightweight model backends against 6 EDR products and 3 payloads, and find that even these simple models generate successful submissions, with DeepSeek-V4.1 Flash reaching 17/18 successful runs and Bitdefender allowing only 5 of 15 runs to pass. These results provide a systematic measurement of current AI agents' endpoint defense evasion capabilities and identify areas for defenders to investigate.

cs.CR

新论文 · v1arXiv:2611.13681

A Functional Preservation Score for Auditing Connectome Operations: Compression, Alignment, Transport, and Local-Rule Writing

Praveg Vashishtha

原文摘要

We introduce the Functional Preservation Score (FPS), a null-normalised, behaviourally grounded measure of how much of a neural network's input-output function survives a structural operation, and use it to audit the operations underlying speculative "mind-migration" pipelines on public data that runs on a single laptop. Across five pre-registered studies on the C. elegans connectome, whole-brain calcium activity, and hippocampal replay, FPS yields two positive characterisations and a three-part negative. Low-bit quantisation of a real connectome exhibits a structure-dependent sparsity floor: uniform int4/ternary quantisation is forced to roughly 73%/98% effective sparsity because heavy-tailed integer synaptic weights round to zero, while mixed-precision protection of hub neurons lowers these floors to about 27%/32%. Decoded hippocampal replay sequences compress heavily yet remain identifiable at a cross-validated top-1 accuracy of 0.69 (95% CI [0.63, 0.74]; chance 0.125; permutation p < 1e-3) from latents of order 10-50 bits per event, and a bounded incremental store shows a graceful-versus-cliff capacity trade-off. The migration pipeline fails in simulation at each stage for largely independent reasons: the putative "identity payload" is not a functionally separable module (p >= 0.16 at every tested size); cross-individual functional transport via Gromov-Wasserstein collapses once the target's wiring is respected (fidelity 1.00 to 0.40, [0.31, 0.49], p < 1e-3) or alignment is imperfect (0.06, [0.00, 0.13]); and biologically plausible local plasticity cannot reach the function the target topology provably admits (gap 0.45, [0.27, 0.64], p < 1e-3, d = 2.0). We release FPS, the data loaders, and all operations as a toolkit. All results are at nematode / small-network scale and concern coherence in simulation, not the feasibility of reading or writing a living brain.

cs.NE · q-bio.NC

新论文 · v1arXiv:2611.13680

Bootstrapped On-Policy Distillation

Yue Wu, Xueyu Zhou, Yangrong Hu, Jian Huang

原文摘要

On-policy distillation (OPD) improves language models by distilling teacher predictions on student-generated trajectories, while on-policy self-distillation (OPSD) removes the need for an external teacher by using a reference-conditioned copy of the model as the teacher. Despite its great early improvements, we find that OPSD does not remain consistently effective under continued training and can eventually suffer substantial performance degradation. We study this training behavior and link the degradation to a growing mismatch between the evolving reference-free student and the frozen privileged teacher. In the early stage of OPSD, the privileged teacher remains closely aligned with the student and provides useful guidance for improving the student's generation trajectories. However, as training progresses, the growing distributional mismatch between the privileged teacher and reference-free student can cause the teacher to provide misleading supervision, ultimately degrading student performance. This phenomenon motivates us to first leverage privileged reference supervision to acquire useful knowledge and then transfer it to a fresh student. We introduce Bootstrapped On-Policy Distillation (BOPD), a two-stage framework that first uses OPSD to internalize task-relevant information from reference-conditioned supervision into an intermediate model, and then distills this reference-free intermediate model into a fresh copy of the initial model. Experiments across models show that BOPD improves mathematical reasoning and code generation beyond both the initial models and the constructed OPSD teachers. These results suggest that privileged reference supervision can be more effectively exploited by first transferring its task-relevant signal into a task-adapted model and then distilling from that model without reference conditioning.

cs.LG · cs.AI

版本更新 · v2arXiv:2611.11166

DexForesight: Grounding Dexterous Policies with Implicit Spatial Foresight

Zhongxi Chen, Yi Yang, Xinyi Chen, Mingzhe Ye, Yaping Li, Zhihong Liu, Congsheng Xu, Haonan Wang 等 9 位作者

原文摘要

Dexterous manipulation requires rich spatial foresight over both the surrounding scene and the robot hand. Existing world-action models often acquire such predictive knowledge through explicit future prediction. We introduce DexForesight, a framework that learns spatial foresight implicitly by injecting dynamics-aware world representations into the latent space of a dexterous policy. During training, DexForesight learns an action-conditioned spatial world model from offline scene depth, kinematic hand geometry, and demonstration trajectories. The world model predicts 3D hand-point motion and visual feature changes, capturing both articulated hand dynamics and scene changes under demonstrated actions, including the motion of fingers occluded from the cameras. We then freeze the world model and align policy queries with its predictive world queries, transferring the learned spatial dynamics into policy latents and using them to condition action generation. At deployment, only the policy is retained, operating on RGB observations, language, and robot state without depth input or explicit future prediction. On 11 DexJoCo simulation tasks, DexForesight improves the overall success rate of the base policy from 49.4% to 60.5%. On three real-robot dexterous manipulation tasks, DexForesight achieves 58.3% success, compared with 50.0% for the strongest baseline. Project page https://dexforesight.github.io/

cs.RO

版本更新 · v2arXiv:2511.21098

Pygmalion Effect in Vision: Image-to-Clay Translation for Reflective Geometry Reconstruction

Gayoung Lee, Junho Kim, Jin-Hwa Kim, Junmo Kim

原文摘要

Understanding reflection remains a long-standing challenge in 3D reconstruction due to the entanglement of appearance and geometry under view-dependent reflections. In this work, we present the Pygmalion Effect in Vision, a novel framework that metaphorically "sculpts" reflective objects into clay-like forms through image-to-clay translation. We then introduce a dual-branch design in which a BRDF-based reflective branch and a clay-guided branch share the same Gaussian geometry but operate through independent rendering paths. The clay-guided branch, supervised by the synthesized clay images, provides reflection-free geometric guidance that complements the photometric supervision of the BRDF branch. Experiments on both synthetic and real datasets show consistent improvements in both geometry and photometric metrics. Beyond technical gains, our framework reveals that seeing by unshining, translating radiance into neutrality, can serve as a powerful inductive bias for reflective object geometry learning.

cs.CV · cs.AI · cs.GR

新论文 · v1arXiv:2611.13677

Deformable Texture Maps: Compiling Dynamic Microgeometry into Neural Textures

Antonio Arian Silaghi, Zhen Chen, Danny Kaufman, Élie Michel, Jean-Marc Thiery, Jorge Lopez-Moreno, Elena Garces

原文摘要

Deformable material surfaces, such as cloth, owe their appearance to fine-scale geometries that restructure under deformation. Even when coarse simulation models efficiently recover this microstructure, explicitly instantiating and rendering it remains expensive. We introduce Deformable Texture Maps (DTM), a representation that compiles deformation-driven microgeometry into a continuous and compact neural texture. Our model maps texture coordinates and local deformations to intrinsic maps describing opacity, surface normals, depth, and material identity. This captures the occlusion and visibility changes generated by deformation. At runtime, these maps provide a common texture-space representation of deformation-driven fine-scale structure that can be evaluated directly on coarse deforming surfaces. Since our representation encodes geometric attributes rather than radiance, lighting and material response remain decoupled from the learned model. This allows us to apply relighting, as well as shading and material editing without retraining. We further introduce a persistent texture-space rendering strategy that caches these intrinsic maps and selectively updates them as deformation changes. This effectively decouples neural evaluation from frame rendering and enables their reuse across frames and conventional rendering effects. We demonstrate our approach on yarn-based textiles, where non-linear material stretch and shear induce changes in yarn geometry, spacing and occlusion. We build our model by, first, sweeping tileable yarn patches through deformation space, then, rendering the detailed geometry into a canonical texture-space, and finally, compressing the resulting maps into a single neural representation. Our resulting model can be evaluated directly on arbitrary coarse garment meshes and integrated with conventional shading pipelines, keeping appearance authorable at render time.

cs.GR

版本更新 · v3arXiv:2512.06493

Programmable GPU-Accelerated dApps for Real-Time Edge Inference and ISAC on NVIDIA Aerial Testbed

Davide Villa, Mauro Belgiovine, Nicholas Hedberg, Michele Polese, Chris Dick, Tommaso Melodia

原文摘要

The transition of cellular networks to (i) software-based systems on commodity hardware and (ii) platforms for services beyond connectivity introduces critical system-level challenges. As sensing emerges as a key feature toward 6G standardization, supporting Integrated Sensing and Communication (ISAC) with limited bandwidth and piggybacking on communication signals, while maintaining high reliability and performance, remains a fundamental challenge. In this paper, we provide two key contributions. First, we present a programmable, open-source framework for processing PHY signals through real-time, GPU-accelerated Artificial Intelligence (AI) applications on the edge Radio Access Network (RAN) infrastructure. Building on the Open RAN dApp architecture, the framework interfaces with a GPU-accelerated gNB based on NVIDIA Aerial Testbed (ATB), feeding PHY data to custom AI logic with about 84 $μ$s of framework latency per slot and control loops with inference in about 123 $μ$s for lightweight TensorRT models, multiple inference engines, and support for several AI backends. We evaluate the framework on two GPU platforms, with different GPU sharing modes and an increasing number of cells and dApps. Second, we demonstrate the framework capabilities through cuSense, an indoor localization dApp that consumes uplink DMRS channel estimates, removes static multipath components, and runs a neural network to infer the position of a moving person. Evaluated on a 3GPP-compliant 5G NR deployment, cuSense achieves a mean localization error of 73 cm on runs unseen during training, with 79% of predictions falling within 1 meter, and runs live in 241 $μ$s, without dedicated sensing hardware or modifications to the 5G signals. The framework is released as open source, together with the cuSense dataset, providing a reference design for future AI-native RANs and ISAC applications.

cs.NI

版本更新 · v2arXiv:2608.08743

A Distribution Mapping Approach to Counterfactually Fair Reinforcement Learning

Jianhan Zhang, Jitao Wang, John D. Piette, Donglin Zeng, Chengchun Shi, Zhenke Wu

原文摘要

Reinforcement learning (RL) seeks to optimize sequential decisions to maximize population-level benefits over time. However, when deployed in high-stakes settings such as healthcare, RL decisions might systematically restrict some subpopulation's access to valuable services in a manner contrary to the values and goals of stakeholders. Counterfactual fairness (CF) offers a promising framework to address this problem based on causal reasoning. This paper develops a data preprocessing algorithm that, when used in tandem with policy learning, enables CF in RL. Our algorithm relies on a novel quantile distribution mapping method for sequentially estimating the counterfactual states and rewards in the data preprocessing step, subsuming common additivity assumptions used for counterfactual prediction as a special case. We theoretically prove that the per-step level of counterfactual unfairness and infinite-horizon suboptimality gap can be bounded under mild regularity conditions. We also empirically test our algorithm in numerical experiments as well as in application to a real-world interventional digital health dataset.

stat.ML · cs.CY · cs.LG · stat.ME

新论文 · v1arXiv:2611.13669

A computational framework for understanding embodied plans

Aalap D. Shah, Basil Saeed, Joshua B. Tenenbaum, Tobias Gerstenberg, Hyowon Gweon, Ilker Yildirim

原文摘要

Reasoning about people's plans and goals based on their actions is fundamental to human social interaction and essential for cooperative, embodied artificial intelligence. Formalizing this capacity in naturalistic, embodied settings has been a longstanding challenge. Here, we present GenEA, a novel agent representation capable of reasoning over multi-step, long-horizon embodied plans. At the heart of GenEA lies a multilevel state-space that couples the coarse biomechanics of an articulated body and its interactions with nearby objects by extending the task-and-motion planning framework from robotics with simulation-based intuitive physics. By embedding this state-space in a generative model, GenEA yields a joint probability distribution over goals and embodied plans. Across several experiments spanning the qualitatively distinct domains of object manipulation and navigation, featuring diverse tasks and multiple behavioral measures, we show that GenEA, through simple transformations of this joint distribution, recapitulates how humans reason about embodied plans. Crucially, GenEA consistently outperforms strong alternative models, including plausible heuristics, standard task-and-motion planning, and a state-of-the-art vision-language model. GenEA thereby offers a new perspective on the architecture of human reasoning and social cognition, suggesting that embodied artificial intelligence systems seeking to seamlessly interact with humans should adopt GenEA-like generative models of action planning.

cs.AI · cs.HC · cs.RO

新论文 · v1arXiv:2611.13667

Benchmarking LLM Next-Step Recommendations with Knowledge Space Theory

Peter Steiner, Jan Hochweber

原文摘要

Selecting an LLM for next-step recommendation requires benchmarks that reflect the intended educational task and domain. We present a benchmarking method that uses Knowledge Space Theory (KST) as a formal evaluation reference. Given a task catalogue and an existing prerequisite relation, the method derives admissible knowledge states and the complete set of immediate next steps for each state. Recommendations generated by a model under a specified prompt are compared with these reference sets. A sample evaluation study uses 2,400 responses from six models across four mathematics domains, four prompt conditions, and two catalogue orders. Exact-set accuracy was 61.3% with descriptions only, 62.8% when models were asked to reconstruct prerequisites, 96.3% with supplied prerequisite relations, and 98.2% with the complete knowledge space. Performance varied across models and domains, and recommendation sets agreed across the two presentations in 85.25% of pairs. The benchmark supports comparisons of model-prompt combinations against a context-specific reference. Its scores describe structural recommendation accuracy under the evaluated conditions; they do not establish learning benefit or the necessity of using an LLM for selection.

cs.AI

版本更新 · v2arXiv:2511.08861

EEG-X: Toward Device-Agnostic and Noise-Robust Foundation Model for EEG

Navid Mohammadi Foumani, Soheila Ghane, Nam Nguyen, Mahsa Salehi, Geoffrey I. Webb, Geoffrey Mackellar

原文摘要

Foundation models for EEG analysis are still in their infancy, limited by two key challenges: (1) variability across datasets caused by differences in recording devices and electrode configurations, and (2) the low signal-to-noise ratio (SNR) of EEG, where neural activity is often buried under artifacts and non-brain sources. To address these challenges, we present EEG-X, a device-agnostic and noise-robust foundation model for EEG representation learning. EEG-X introduces a location-based channel embedding that enables robust transfer across heterogeneous EEG devices and electrode layouts. To improve robustness against noise, EEG-X employs a novel noise-aware masking-reconstruction strategy that reconstructs artifact-cleaned signals rather than raw noisy EEG and introduces a novel Dictionary-based Convolutional Transformation (DiCT) that improves reconstruction-based pretraining by comparing signals in a structured feature space instead of directly in the raw signal space. Experiments across datasets collected from diverse EEG devices show that EEG-X consistently outperforms state-of-the-art methods across multiple downstream tasks and demonstrates strong cross-domain generalization when pretraining and downstream datasets differ in electrode layouts and recording configurations paving the way towards foundation models. The models and code are available at https://github.com/Emotiv/EEG-X.

cs.LG · cs.HC

新论文 · v1arXiv:2611.13662

NIF-VIO: Visual--Inertial Odometry Across Inertial and Non-Inertial Reference Frames

Zichao Shu, Lijun Li, Zhuochen Li, Zetao Chen, Jianyu Wang

原文摘要

Conventional visual--inertial odometry (VIO) assumes that observations from cameras and an inertial measurement unit (IMU) are jointly consistent in an inertial reference frame. However, this assumption breaks down under non-inertial conditions. Inside accelerating or rotating carriers, visual observations of the interior reflect carrier-relative motion, while IMU measurements contain additional carrier-induced components, which can compromise conventional fusion. We therefore present NIF-VIO for localization relative to the local environment across inertial and non-inertial reference frames. It adapts visual--inertial fusion based on measurement consistency. As inertial conditions change, early constraint isolation is combined with local reoptimization to limit error propagation before inconsistency confirmation and repair the affected local geometry. Under non-inertial conditions, a compact model with temporally shared parameters is introduced to account for carrier motion while preserving inertial information that complements visual constraints. Experiments on public benchmarks and a self-collected dataset demonstrate that NIF-VIO achieves competitive performance under inertial conditions and improved localization accuracy and availability under non-inertial conditions.

cs.RO