跳到正文
研踪PaperTrail

YOUR RESEARCH, IN SIGHT

跟进版本变化

查看已发布论文的最新修订，在详情页对照实际观测到的版本。

部分采集异常

全站更新概况

+1673新增论文
+1257其中 AI 相关
+198已有论文修订
+100新增 AI 导读

arXiv 最近完成扫描：2026-11-14 02:49 · 全站最近入库：2026-11-13 20:44各信源进度 →

统计口径与数据新鲜度

统计 2026-11-13 03:19 — 2026-11-14 03:19（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：2026-11-14 02:49最近发现新增版本：2026-11-13 20:44库内论文最新更新时间：2026-11-13 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。

23,348 已收录论文19,580 AI 相关6,727 有修订112 HF 社区发现

论文修订列表

研究日报 / 周报 →按信源浏览 →

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 6,727 篇论文 · 按版本更新时间排序上轮扫描起点 2026-11-14 02:48（北京时间）
版本更新 · v3arXiv:2611.05397

Ranking and Unranking Permutations in Heap's Order

Yusheng Hu

原文摘要

Heap's algorithm is a permutation generation method with very low constant overhead. Its element-interchanging mechanism, however, prevents the standard factorial-base Lehmer code from being applied directly, and to our knowledge no explicit Rank/Unrank algorithm for the Heap order has been published. This paper describes the block structure of Heap's algorithm in terms of rearrangement operators: the net effect of a block-internal traversal is a tail-preserving operator and the block-boundary swap is a top-level swap operator, and together they yield a closed-form recurrence for the block transitions. On this basis a bottom-up Unrank algorithm and a top-down Rank algorithm are built from one and the same closed-form in-block mapping; both run in $O(n^2)$ time and $O(n)$ space and require no factorial-scale precomputed table. The supporting proofs use a synchronized strong induction on the dimension to establish the bijectivity of the trailing-element mapping, closed forms for the final arrangement and for the trailing sequence, the dichotomy of its cycle structure, and the reciprocity of Rank and Unrank. The Heap order is in fact the colex order of the mixed-radix code.

cs.DS

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2604.22318

Strategically Robust Linear Quadratic Dynamic Games

Boris Velasevic, Nicolas Lanzetti, Eric Mazumdar

原文摘要

We study linear quadratic dynamic games where players are uncertain about each other's control policies or goals and consequently seek to be strategically robust. Building on recent work on strategically robust and risk-averse game theory, we first formalize the problem of strategically robust linear quadratic dynamic games. We show that these can be rewritten as simple transformations of linear quadratic games in which each player chooses a controller in a fictitious game in which they are faced with an adversary who is penalized for deviating from the other players' policies. This formulation naturally induces a novel notion of dynamic equilibrium, which we call a strategically robust dynamic equilibrium. We establish existence and uniqueness (in the subgame-perfect sense) of such equilibria and show that the equilibrium policies are Markovian, linear, and can be efficiently computed via coupled backward Riccati equations. Through numerical simulations, including experiments in a network game, we illustrate the benefits of strategic robustness in designing robust and resilient decentralized control schemes. Our experiments also expose a "free-lunch" phenomenon in games in which robustness does not incur a corresponding loss in performance but can yield improvements in players' utilities and social welfare.

math.OC · cs.GT · eess.SY

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2607.21776

PulseGuard: rPPG-Derived Temporal Signals as a Forensic Modality for Talking-Face Deepfake Detection

Othmane Harraq, Tamer Aldwairi

原文摘要

Talking-face (TF) deepfake generation synthesizes photorealistic facial video from a static source image and an audio signal, producing forgeries that current image-based detectors consistently fail to identify. Unlike face-swap manipulation, TF synthesis has no underlying real video from which to inherit physiological characteristics, motivating remote photoplethysmography (rPPG) as a detection modality for this forgery category. We propose PulseGuard, a detection framework that extracts per-window rPPG waveforms via RhythmFormer and trains a suite of lightweight classifiers to distinguish real from synthesized traces. Evaluated on the TF subset of Celeb-DF++ under a strict subject-independent protocol, where test identities are completely separated from training identities, our 1D ResNet achieves a window-level AUC of 0.826 at EER 25.1% (0.842 pooled to video level, the granularity at which DeepFakesON-Phys is scored) while operating exclusively on the extracted temporal signal, and retains AUC 0.754 on generators withheld entirely from training. We document a controlled reproduction study of DeepFakesON-Phys, the representative prior rPPG detector, demonstrating degradation from AUC 0.999 on legacy face-swap data to 0.576 on the TF subset of Celeb-DF++. We further show that detection difficulty is strongly method-dependent: AUC ranges from 0.937 (Real3DPortrait) to 0.617 (IP-LAP) across the seven TF generators, with the ordering broadly preserved under per-method training. This spread is stable across training regimes rather than reflecting evaluation noise, and constitutes the primary empirical contribution of the work. We report these as empirical baselines and do not establish that the discriminative content is cardiac in origin.

cs.LG · cs.CR · cs.CV · cs.MM

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2605.07755

Rethinking State Tracking in Recurrent Models Through Error Control Dynamics

Jiwan Chung, Heechan Choi, Seon Joo Kim

原文摘要

The theory of state tracking in recurrent architectures has predominantly focused on expressive capacity: whether a fixed architecture can theoretically realize a set of symbolic transition rules. We argue that equally important is error control, the dynamics governing hidden-state drift along the directions that distinguish symbolic states. We prove that affine recurrent networks, a class of models encompassing State-Space Models and Linear Attention, cannot correct errors along state-separating subspaces once they preserve state representations. Consequently, practical affine trackers do not learn robust state tracking; rather, they learn finite-horizon solutions governed by accumulated state-relevant error. We characterize the mechanics of this failure, showing that tracking remains readable only while the accumulating within-class spread remains small relative to the between-class separation. We demonstrate empirically on group state-tracking tasks that this breakdown is predictable: tracking collapses when the distinguishability ratio crosses the readability threshold. Across trained models, the point of this crossing predicts the horizon at which downstream accuracy fails. These results establish that robust state tracking is determined not only by an architecture's theoretical expressivity but crucially by its error control.

cs.LG · cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.10453

BunshinRL: Whole-Workspace Forking for Agentic Reinforcement Learning Rollouts

Yutong Huang, Vikranth Srivatsa, Alex Asch, Yiying Zhang

原文摘要

Agentic reinforcement learning (RL) rollouts interleave model reasoning with tool calls in persistent workspaces. Scaling these rollouts incurs substantial memory costs for concurrent environments, while slow tool calls delay entire rollout batches. We present BunshinRL, an agentic RL rollout system that exploits shared environment state and recurring tool calls through three techniques: memory sharing across environments, tool-prefix caching, and speculative tool execution. These techniques build on BunshinFork, a novel live whole-workspace forking mechanism for containers. BunshinFork reconstructs process trees, shares memory copy-on-write, and preserves file, communication, and GUI state, enabling cached-state reuse and rollback after incorrect speculation. We implement BunshinRL and BunshinFork through Linux kernel module and CRIU modifications. BunshinRL reduces peak memory by up to 43%, total memory consumption by up to 51%, and end-to-end rollout latency by 23%.

cs.AI · cs.OS

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.16581

EmoPhone: A Multi-Wave Dataset for In-the-Wild Mobile and Wearable Affect Sensing

Panyu Zhang, Minseo Park, Soowon Kang, Tomiris Ismatzoda, Azizbek Mustafakulov, Otabek Najimov, Woohyeok Choi, Jumabek Alikhanov 等 10 位作者

原文摘要

We introduce a three-wave, in-the-wild multimodal dataset for affect sensing that integrates smartphone sensing, wearable sensing, and dense experience-sampling-method (ESM) labels collected annually from 2020 to 2022. The dataset supports moment-level affect modeling through a shared dimensional label core across all waves, with additional affective descriptors available in the third wave (D-3). We describe the resource in terms of study design, temporal density of in-situ labels, and sensing and label coverage across waves. To support evaluation within this resource, we define an initial three-setting benchmark spanning temporal prediction from within-user history, within-wave cross-user generalization, and cross-wave generalization in which each wave is treated as a separate dataset. Our benchmark results show that the strongest method family depends on the evaluation setting: supervised baselines perform best in the temporal setting, unsupervised domain adaptation is strongest overall in the within-wave cross-user setting, and domain generalization shows the strongest overall cross-wave performance, although its margin over strong baselines is modest. These findings indicate that robust mobile affective computing is constrained not only by label availability but also by substantial participant-level variability and realistic cross-wave differences inherent in longitudinal in-situ deployments.

cs.HC

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.13755

A Hardware-Trained Generative Quantum Eigensolver for Sample-Based Quantum Diagonalization of Organotin Hydroxides up to 70 Qubits

Kunal Kumar, Kimberlee Keithley, Toshiya Iwamori, Shu Kanno, Ikko Hamamura, Takao Kobayashi, Akira Koseki, Qi Gao 等 15 位作者

原文摘要

Quantum hardware can provide feedback for training circuit generators beyond the reach of exact classical simulation. We train generative quantum eigensolver (GQE) policies, transformers conditioned on molecular descriptors, to generate the state-preparation circuits for sample-based quantum diagonalization (SQD). Training runs first on a noiseless simulator and then with the IBM Heron processor ibm_kobe inside the loop, at 46 and 70 qubits ($1.2 \times 10^8$ shots for the four main training runs). A classical CCSD(T) reference energy of the training molecules enters only an admission gate and two auxiliary reward terms. During hardware training, a reward penalty prevents collapse to circuits that sample only the Hartree-Fock determinant. Policies trained on transition-metal complexes are applied without retraining to Sn(OH)$_3$-Me and its hydrates, minimal models of organotin photoresists for extreme-ultraviolet lithography, from 32 to 70 qubits. After one round of hardware sampling, followed by classical ranking and recovery of the cached counts with the molecular Hamiltonian and no further QPU calls, 27 of 28 runs at 46-70 qubits are within 15 mHa of the density-matrix renormalization group (DMRG) reference (median 4.5 mHa), and in all eight runs with an untrained control the trained policy is more accurate. At 56-70 qubits the selected circuits use 34-72 times fewer two-qubit gates than the local unitary cluster Jastrow (LUCJ) circuit transpiled for the same device. These results demonstrate that circuit generators can be trained with quantum hardware at 70 qubits.

quant-ph · cs.AI · physics.chem-ph

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.13885

Are Causal Neuron Circuits Unique and Transferable?

Ömer Tekin Yavuz

原文摘要

Interpretability demonstrations often show that intervening on a group of neurons changes a model's answer. A recent example removes an 18-neuron group in Qwen3-32B and the model answers "Simba" instead of "Scar" to a two-hop Lion King riddle. Such a demonstration establishes that the group matters, but not how much of it is necessary, whether other groups could do the same, or whether the effect carries over to other inputs. We ask: when intervening on a group of neurons changes a model's answer, how much of that group is actually necessary? For every selected group we measure all 4,096 subsets of 12 MLP neurons. On a two-alternative capital task (Qwen3-0.6B), zero-ablating 12 selected neurons lowers the answer margin by 27% while random neurons barely change it, yet replacing the same neurons with their values from other questions also barely changes it: their activations are nearly constant across questions, and we find no evidence that they carry question-specific content. On two-hop relation chains, patching 12 neurons with activations from the first-hop question shifts the answer to the bridge entity (from Scar to Simba): the greedy answer starts with the bridge in 9/11 chains for Qwen3-1.7B, 9/11 for Llama-3.2-1B and 8/10 for Gemma-3-1B; random neurons never do. The full group is rarely necessary: a few neurons often suffice, and in many chains disjoint subsets each flip the answer on their own, in an exploratory analysis most often when the model is less confident. Group effects transfer to rephrasings of the same chain but are much weaker on other tested chains. We study neuron-level causal interventions and alternative subsets; we do not recover the full connectivity of circuits.

cs.LG · cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.12870

SWAM: Scaling World Action Models with Cross-Modal Alignment

Yu Zhou, Xinyuan Xia, Xuankun Yang, Yinghao Yang, Wendi Chen, Bowen Liu, Jianhan Ma, Weicheng Duan 等 11 位作者

原文摘要

World action models jointly predict video and actions conditioned on language. Existing studies focus primarily on architectures and training mechanisms, typically treating language-video alignment acquired through video pretraining as an established foundation. However, introducing action learning may impair this alignment. Principled guidance on data and supervision design for maintaining language-video alignment during action learning remains limited. To investigate this, we curate a pool of approximately 47,000 hours of general videos, egocentric videos, UMI demonstrations, and robot demonstrations. Building on this, we systematically study how language-video supervision, language-video-action supervision, and the joint use of these two forms affect WAM performance. In particular, we examine how language annotation quality affects pretrained language-video capabilities, and how action supervision from UMI and robot demonstrations extends these capabilities to language-video-action modeling. To better scale the model, we further conduct detailed studies on joint training recipe and model architecture. Under a fixed training budget, mixing egocentric videos and UMI demonstrations improves control generalization. Model architecture affects how effectively these experiences are used, while scaling pretraining under the selected configuration brings further gains. The resulting recipe, post-trained on RoboTwin Clean demonstrations only, achieves success rates of 88.10% and 61.56% under Easy (Clean-to-Clean) and Hard (Clean-to-Random) conditions, respectively. Our model achieves 40.00% SR across three real-world tasks, compared with 23.33% for the strongest baseline, indicating better generalization under unseen conditions. We release the curated data, training code, and model weights.

cs.CV · cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2605.05659

Structural Correspondence and Universal Approximation in Diagonal plus Low-Rank Neural Networks

Ying Chen, Aoxi Li, Jihun Kim, Javad Lavaei

原文摘要

Structured low-rank and sparse-plus-low-rank parameterizations, motivated by methods such as LoRA and low-rank factorization, impose strong structural constraints on the linear transformations within neural networks. We study the expressive consequences of these constrained structures from the perspective of universal approximation. We first characterize that purely low-rank neural networks, despite being able to exactly interpolate arbitrary finite scalar datasets, can fail to approximate continuous functions that vary along directions ignored by the low-rank hidden mapping. We show that augmenting the low-rank structure with a one-parameter diagonal component suffices to overcome this geometric bottleneck, yielding a Diagonal-plus-Low-Rank (DLoR) structure. We show that a full-rank linear transformation can be reconstructed using DLoR components either by expanding network width through additive decomposition or by extending network depth through multiplicative decomposition. We term this correspondence between additive width expansion and multiplicative depth extension the Structural Correspondence framework. By combining these constructions with local linearization of the activation function, we prove that DLoR neural networks restore universal approximation for general activation functions. These results characterize how architectural width and depth can compensate for severe rank constraints without requiring unrestricted dense hidden transformations. Diagonal terms are required for multiplicative construction, while additive expansion works with purely low-rank terms. We further determine its exact cost: both the minimal number of rank-$r$ DLoR factors representing $W$ and the minimal network depth for that layer equal $\lceilκ(W)/r\rceil$, where $κ(W)=\min_{γ\ne0}rank(W-γI)$, achieving a lower cost than its dense representation when $W$ is identity-centered.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.34098

Beyond Correctness: Evaluating Semantic Knowledge in Cross-Table Transfer

Seokyong Sheem, Hochang Lee, Suyeong Lee, Daekyum Kim

原文摘要

Semantic knowledge is increasingly used to bridge heterogeneous schemas in tabular learning, but how much does that knowledge actually improve prediction? Studies in tabular learning commonly answer this question through semantic ablations that modify or suppress the supplied semantic knowledge. We show that these ablations can lead to misleading conclusions about predictive benefit: poor performance under altered semantics may be taken as evidence that the intended knowledge is beneficial. Across real and controlled experiments, altering semantic content can produce large performance differences even when the model gains little predictive benefit from having that semantic knowledge in the first place. To separate these effects, we distinguish two quantities: content sensitivity and predictive utility. Content sensitivity measures the change in performance when semantic content is altered, whereas predictive utility measures the benefit of the intended semantic knowledge relative to a suitable reference without that knowledge. This distinction motivates an evaluation framework in which the control is chosen according to the question being asked: altered controls assess sensitivity to semantic content, whereas claims that semantic knowledge improves prediction require a suitable reference. Even then, predictive utility is not fixed; it varies across suitable references and decreases when the reference can more easily recover the tested knowledge from other inputs or labeled examples. In a bounded audit of 25 semantic-ablation comparisons across nine studies, only one of 18 explicit predictive-utility claims is paired with a control that clearly isolates the tested semantic contribution. Together, these findings motivate a simple evaluation principle: semantic-ablation controls should be chosen and interpreted according to the question they are intended to answer.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.10555

YRC-Bench: Benchmarking Culturally Aware Machine Translation for Historical Heritage via Tri-Level Semantic Bridging

Yan Huang, Aobo Li, Jiayi Wang, Huidong Zhu

原文摘要

Translating historical cultural heritage poses profound challenges for machine translation systems due to the acute semantic asymmetry across eras and cultures, where terse classical syntax frequently obscures domain-specific cultural knowledge. While Large Language Models have demonstrated remarkable prowess in generic translation benchmarks, they struggle to negotiate Culturally-Specific Items and diachronic shifts in domain-rich historical texts, such as Yellow River cultural records. To bridge this critical gap, we introduce YRC-Bench, the first human-curated, multi-parallel benchmark dedicated to historical heritage translation across Classical Chinese, Modern Chinese, and nine target foreign languages. YRC-Bench establishes a tri-level parallel corpus incorporating modern explanatory glosses as explicit semantic bridges and operationalizes translation strategies through paired conserved versus substituted CSI variants categorized by cultural translatability. Systematic evaluations of nine leading LLMs across bidirectional translation directions reveal that leveraging modern Chinese glosses - especially human-curated ones - dramatically elevates both lexical fidelity and semantic adequacy. Furthermore, multidimensional human preference audits by native experts unveil distinct cross-cultural preferences between Sinographic-tradition and non-Sinographic language groups.

cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.11975

AutoQS: Automated Quantization Strategy Design with Large Language Models

Zhen Liu, Zhiqin Yang, Yonggang Zhang, Jianyi Liu, Jinjun Wang, Jingwen Fu

原文摘要

Quantization reduces LLM deployment costs, but effective strategies must balance model quality and storage. LLM-driven search can automate strategy design through repeated proposal and evaluation. This process must identify useful strategies within a limited evaluation budget. The key challenge is that \textit{actual quantization is expensive, while inexpensive proxies may not reflect quantized-model performance.} To address this challenge, we present AutoQS, an \textbf{Auto}mated \textbf{Q}uantization \textbf{S}trategy Design with LLMs. AutoQS keeps weight precision fixed and searches group-size allocations across layers and operators under deployment constraints. Evaluation history guides coordinated LLM proposals, while deterministic validation ensures compatible allocations. An inexpensive quality proxy evaluates candidates, and an candidate repository retains alternative allocations to inform subsequent proposals. Periodic actual quantization complements these proxy evaluations with measured quality and checkpoint storage, providing feedback for further search within the evaluation budget. Experiments on Qwen2.5-7B-Instruct show that AutoQS improves seven-task average accuracy over uniform-group GPTQ and AWQ by 0.70 and 1.12 percentage points, respectively, at four-bit weight precision.

cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.14505

Continual-ARC: Measuring Skill Acquisition over a Working Lifetime

Daniel Lenton, Yusha Arif, Haris Mahmood

原文摘要

Agent harnesses have emerged as a promising mechanism for continual learning, alongside test-time parametric learning. However, the current benchmarks have several limitations: they generally focus on a single applied domain, serve each instance only once, and assume a single family of learner, so memory-based and weight-based systems are rarely compared on the same scale. We propose Continual-ARC, a lifelong stream built from the three ARC-AGI editions, whose tasks need no domain knowledge. The same tasks return after dormant gaps with fresh instances, while new tasks keep arriving. On each instance, the learner may pay for demonstrations, submit answers, or act inside the harness, with retries permitted on incorrect submissions, which are paid for. Each instance is scored based on the share of this feedback budget consumed. Since a learner may keep anything between instances, memory-based and weight-based systems are scored on the same scale. After the stream, the learner is frozen and tested on fresh instances of the tasks it met, which measures memorization, and on unseen tasks, which measures generalization. We run an agent harness with frozen weights, with and without persistence, on Qwen3.5-9B and GLM-5.3 Flash, comparing them with two prompting floors, a program-induction method and Qwen3.5-9B's weights trained to imitate GLM-5.3 Flash. Persistence in the harness generally does not improve memorization unless provided with explicit expert guidance. Most harness-based methods still learn less than a floor that simply caches each task's demonstrations, across all models we tested. No harness-based method gains on unseen tasks, whereas weight-based learning shows substantial gains. This opens the door to future weight-based continual learning that could teach models to drive harnesses in new domains.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2609.11174

A Four-Valued Graph Model for Conflict Resolution: Core Framework and a Machine-Checked Formalization in Lean 4

Yukiko Kato

原文摘要

This note consolidates the core of the Quasi-Closed World Graph Model for Conflict Resolution (QCW-GMCR), which extends the standard Graph Model for Conflict Resolution with Belnap's four-valued logic to represent option-level epistemic ambiguity, and pairs the framework with a machine-checked Lean 4 formalization. QCW-GMCR combines: (1) FOUR-valued option assignments with compositional propagation to state-level feasibility; (2) graded reachability (definite, credible, possible) based on an FDE-inspired transition-warrant semantics, with definite reachability related to FDE consequence in the Boolean fragment; (3) axiomatized deterministic reductions from four-valued assessments to binary decisions, including four canonical operators reflecting distinct risk attitudes; and (4) catastrophe-avoiding equilibrium concepts with a quasi-closed-world safety invariant. A four-valued hypergame extension captures heterogeneous subjective assessments across decision makers. We state the core definitions and results and report the parts verified in Lean 4 with mathlib, including the classical GMCR stability hierarchy, algebraic and compositional properties of FOUR-valued conjunction, properties of the canonical reductions, and the graded reachability hierarchy. The formalization also helped identify and correct earlier claims, including a knowledge-monotonicity axiom replaced by truth monotonicity. This preprint provides a stable, citable record of the framework and its current formal verification status.

cs.LO

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2606.08691

Hierarchical Projection for Adaptive Knowledge Transfer

Samhita Pal, Tian Gu

原文摘要

Transfer learning aims to improve estimation in a target population with limited data by borrowing information from related external datasets, referred to as sources. However, heterogeneity in target-source relevance and predictor-outcome relationships can lead to negative transfer, where model performance is worse than using target data alone, particularly in high-dimensional settings with partially overlapping features. We propose Projection Transfer Learning (ProjectionTL), a Bayesian framework that integrates hierarchical modeling with adaptive posterior projection to enable selective knowledge transfer at both the source and feature levels. In the first stage, a source-guided hierarchical prior adaptively combines information from multiple sources using data-driven weights that capture their global alignment with the target. In the second stage, a weighted posterior-projection procedure induces sparsity by preferentially retaining features supported by stronger agreement between source and target signals. This two-stage framework distinguishes global source relevance from coordinate-specific transferability, allowing informative components to be retained while reducing the influence of heterogeneous or non-transferable signals. We establish theoretical guarantees for posterior contraction and model selection consistency under high-dimensional regularity conditions, without requiring all sources to be informative. Simulation studies and applications to Alzheimer's disease neuroimaging data demonstrate improved estimation accuracy, variable selection, and predictive performance compared with others, including settings with heterogeneous sources and block-wise missing modalities.

cs.LG · stat.ME

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.36794

RAE-PPG: Duration-Grounded Retain-and-Extend Pretraining for PPG Foundation Models

Suyeong Lee, Hochang Lee, Seokyong Sheem, Daekyum Kim

原文摘要

Signal features derived from photoplethysmography (PPG) require different signal durations to characterize. Existing PPG foundation models treat duration as a pretraining or evaluation condition rather than using the different durations required by PPG features to organize self-supervision. We hypothesize that self-supervision should expand with signal duration, allowing a single encoder to progressively acquire additional features while preserving and reusing earlier learning. We introduce Retain-and-Extend PPG (RAE-PPG), which trains a single Transformer encoder successively on 10 s, 30 s, and 240 s inputs, adding supervision for signal features supported by each longer observation. The encoder is partitioned into duration-specific parameter groups, allowing later stages to reuse earlier groups while updating only the group assigned to the current stage. Selected earlier targets are reused to supervise later stages, encouraging the corresponding features to remain accessible in longer-input representations. Direct decoding from the final encoder shows that earlier features remain recoverable from longer-input representations, while later-stage features show higher mean decoding performance at their introduction durations. Controlled comparisons further show that prior-stage learning provides a better basis for learning newly introduced features at both transitions. Across 18 tasks from eight datasets, the final frozen encoder achieves the best observed score on 12 tasks compared with five existing PPG foundation models.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2610.41354

OneSteer: Unlocking Steerable General Humanoid Motion Tracking via One-Step Diffusion Whole-Body Control

Yifei Huang, Yazhao Yang, Li Weijia, Tianle Shi, Ji Jin, Zhihui Cai, Zihao Liu, Jianzhuang Zhao 等 13 位作者

原文摘要

Deployment objectives often specify desired outcomes without providing revised trajectories, while parts of the original reference must still be followed. We present OneSteer, a one-step diffusion whole-body controller that combines general motion tracking with selective test-time steering. To accommodate new objectives, differentiable task costs guide optimization of the generator's input noise with frozen weights, while body-group reference masks relax selected tracking requirements. Predicted state-action plans provide a shared interface for task optimization and feedback-conditioned execution. For nominal tracking, the model trained from scratch achieves the lowest local tracking error among the evaluated baselines on successful held-out AMASS and BONES clips. For steering, one frozen distilled controller adapts across diverse deployment objectives in simulation, raising the clean passage rate under a 1.25 m bar from 2% to 90%. On a physical Unitree G1, OneSteer completes all 27 nominal tracking trials across nine diverse whole-body motion sequences without falling. With a steering cost derived from a signed distance field built using only onboard LiDAR, it also passes under a 1.1 m low bar in four of five attempts and steps over a 12 cm hurdle without falling in all three attempts. Beyond control, prediction variation in the model trained from scratch ranks likely tracking failures without physics rollouts. This scoring method is more than 10x faster than simulation-based scoring with five rollouts per clip and guides targeted fine-tuning. Project page: https://onesteer.github.io

cs.RO

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.06658

The determinantal complexity of the 4*4 permanent is 15

Karthik Sheshadri

原文摘要

The best known lower bound on the determinantal complexity of the 4x4 permanent is 10. In this short note, I prove that this value is exactly 15: the 4x4 permanent is the determinant of a 15x15 matrix of affine linear forms, and cannot be that of any smaller matrix. To do so, I must establish both a lower and upper bound. To establish the upper bound, I use Grenet's construction of size 2^n-1. Hence that construction is optimal at n=4 (and n=3). This now constitutes the first knwon exact value of the determinantal complexity of the permanent at n=4. Only the value at n=3 was exactly known, prior to this note. The general quadratic bound only gives 8 for n=4. To establish the lower bound, I work from a normal form at a zero of the permanent. In the first row (and column) of the normal form, the linear forms vanish on a linear space inside the singular locus of the permanent. A hessian inequality bounds the size of the matrix in terms of that space. Then, a matrix of sie 14 foces the space into any of nine corordinate types. The forcing is either up to symmetry, or else into a family that degenerates to one of them. I then rule out every case. All involved computations are exact. In respect of the claim, I have made all certificates public.

cs.CC

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2511.23115

Robust Image Emotion Classification via Multi-Level Affective Captioning

Zibo Zhou, Zhengjun Zhai, Huimin Chen

原文摘要

Image emotion classification (IEC) has been studied for many years and has recently attracted increasing attention with the rapid development of deep learning. Although recent studies have benefited from knowledge embedded in pre-trained visual models, the performance is still limited by the "affective gap" between low-level visual features and high-level affective semantics. Psychological studies suggest that language can convey fine-grained and emotion-rich information. Inspired by this observation, we attempt to narrow the "affective gap" by introducing affective text. Specifically, we design a multi-level contrastive loss to detect emotional concepts from images and introduce an emotional-attribute-guided large language model approach to generate affective sentences. Then, a pre-trained language model is employed to combine emotional concepts with affective sentences for IEC. In addition, we introduce a contrastive loss with a semantic-similarity-based sampling strategy to address large intra-class differences and limited inter-class differences in affective datasets. Moreover, unlike many previous studies, we explicitly consider images containing embedded text. Extensive experiments on multiple benchmarks show that the proposed method helps narrow the "affective gap" and achieves competitive performance.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2603.22570

CanViT: Toward Active-Vision Foundation Models

Yohaï-Eliel Berreby, Sabrina Du, Audrey Durand, B. Suresh Krishna

原文摘要

Active computer vision promises efficient, biologically plausible perception through sequential, localized glimpses, but lacks scalable general-purpose architectures and pretraining pipelines, leaving Active-Vision Foundation Models (AVFMs) underexplored. We introduce the Canvas Vision Transformer (CanViT), the first task- and policy-agnostic AVFM. CanViT uses scene-relative RoPE to bind a retinotopic Vision Transformer backbone and a spatiotopic scene-wide latent workspace, the canvas. Efficient interaction with this high-capacity working memory is supported by Canvas Attention, a novel asymmetric cross-attention mechanism. We decouple thinking (backbone-level) and memory (canvas-level), eliminating canvas-side self-attention and fully-connected layers to achieve fast sequential inference and scalability to high output resolutions. We propose a label-free active vision pretraining scheme, policy-agnostic passive-to-active dense latent distillation: reconstructing scene-wide DINOv3 embeddings from sequences of low-resolution glimpses with randomized locations, zoom levels, and lengths. We pretrain CanViT-B from a random initialization on 13.2 million ImageNet-21k scenes and 1 billion random glimpses, in 166 hours on a single H100. On ADE20K segmentation, a frozen CanViT-B achieves 38.5% mIoU in a single low-resolution glimpse, outperforming the best active model's 27.6% with 20x fewer inference FLOPs as well as its FLOP- or input-matched DINOv3 teacher. Given additional glimpses, CanViT-B reaches 45.9% ADE20K mIoU. On ImageNet-1k classification, CanViT-B also sets a new active-vision state of the art, with 84.5% top-1 accuracy after fine-tuning. CanViT generalizes to longer rollouts, larger scenes, and new policies. Our work narrows the wide gap between passive and active computer vision, demonstrating the potential of task- and policy-agnostic AVFM pretraining.

cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.08714

Patching the Blind Spots: Reversible LLM Rule Overlay for Frozen Fraud Detectors

Bo Qu, Zhurong Wang, Frank Zahradnik, Ming Ouyang

原文摘要

Gradient-boosted trees dominate industrial tabular fraud classification, but even a well-trained model has blind spots, and retraining a deployed model is not always safe. We bring LLM reasoning to patching such a model without modifying it: LLMs read its tree digest and SHAP attributions, reason about the fraud it misses, and propose rules anchored on features it underweights, screened and calibrated into a reversible log-lift overlay at the production cut-off. Single-LLM benefit is high-variance and inverts between datasets, so we pool three diverse backbones by logit averaging, trading peak gain for a bounded worst case. On an experimental detector trained on sampled, selection-biased real-world e-commerce data, tuned continued training regresses, whereas the overlay gains on average +2.26% recall at equal precision ($Δ$R@P) and +1.60% precision at equal recall ($Δ$P@R). Replayed offline on a legacy production detector, the overlay, calibrated on one month and frozen, stays more precise at equal recall for seven months (+5.64% $Δ$P@R). On seven bias-shift variants of a public benchmark it raises $Δ$P@R on all and keeps $Δ$R@P near neutral. LLM reasoning over a model's diagnostics can improve a deployed tabular classifier through readable predicates naming the fraud they cover.

cs.LG · cs.AI · cs.CL

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2609.10255

Sharp Bounds on the Number of Small Cuts

Yu Cong, Chao Xu

原文摘要

A $k$-cut of an undirected graph is a minimal edge set whose removal leaves at least $k$ connected components, and its cost is the sum of its edge costs. Write $λ_k$ for the minimum cost of a $k$-cut. For fixed $k\ge2$ and $α\ge1$, we show that in any $n$-vertex graph with $n\ge k$ and nonnegative edge costs, the number of $k$-cuts of cost strictly below $αλ_k$ is $O(n^{\lceilαk\rceil-1})$. This implies an $O(n^{\lfloorαk\rfloor})$ bound for $k$-cuts of cost at most $αλ_k$, sharpening the $O(n^{αk})$ bound of Gupta, Harris, Lee and Li. For $k=2$, our strict bound resolves a conjecture of Henzinger and Williamson and an open question of Karger. The exponent $\lceilαk\rceil-1$ is optimal for $k$-cuts for every $k\ge2$ and $α>1$.

math.CO · cs.DM

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.13219

MediPFN: Amortized In-Context Nuisance Learning and Inference for Causal Mediation Analysis

Shi Bo, Jiaming Hu

原文摘要

Causal mediation analysis decomposes a treatment effect into indirect pathways through mediators and direct pathways through other processes. Tabular foundation models perform classification and regression through in-context learning, providing estimates of conditional means and propensity scores used in causal mediation analysis. Constructing confidence intervals from these estimates requires accounting for uncertainty introduced by fitting and combining several nuisance functions. We introduce MediPFN, an amortized inference framework that adapts a pretrained tabular foundation model to nuisance prediction and combines its predictions through a cross-fitted influence-function estimator. A separate head learns quantiles of the estimator's standardized error from simulated tables with known indirect effects. Supervised fine-tuning, symmetrization, and independent calibration then produce a confidence interval. We establish coverage bounds for the resulting intervals under standard assumptions. Across experiments, MediPFN attains coverage closer to the nominal level than several comparators. It also reduces the computational cost of uncertainty quantification, running 59.1 times faster than bootstrap in matched comparisons.

stat.ML · cs.LG · stat.ME

查看论文
arXiv ↗PDF ↗
版本更新 · v3arXiv:2606.09547

Streaming Interventions: Can Video Large Language Models Correct Mistakes as They Occur?

Apratim Bhattacharyya, Shweta Mahajan, Sanjay Haresh, Rajeev Yasarla, Reza Pourreza, Litian Liu, Risheek Garrepalli, Roland Memisevic

原文摘要

Learning everyday skills, like cooking a dish, relies increasingly on instructional media such as online videos. This opens the door to the use of video (and multimodal) large language models (LLMs) as task guidance assistants. A crucial capability for the real-world success of a prospective task guidance assistant is it's ability to intervene proactively as soon as a mistake is apparent in order to guide the user. To evaluate this crucial capability, we introduce Ego-MC-Bench (Mistake Corrections), a benchmark for evaluating reactive, step-by-step task guidance in realistic cooking scenarios. Extensive experiments show that Ego-MC-Bench is highly challenging for state-of-the-art video LLMs. We argue that a key reason is the limited availability of training data for fine-tuning models on this task. Although there exists a wide range of cooking video datasets, existing datasets lack examples of mistakes along with appropriately timed interventions. To help address this data limitation, we also introduce Ego-CoMist, a counterfactual synthetic dataset created by transforming non -interactive cooking videos into supervised training examples showing proactive interventions. We show that fine-tuning on Ego-CoMist yields performance gains especially for smaller and more efficient video LLMs that are well suited for delivering assistance on edge devices.

cs.CV · cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2603.00397

TENG-BC: Time-Evolving Natural Gradient for High-Accuracy Neural PDE Solvers with General Boundary Conditions

Hongjie Jiang, Di Luo

原文摘要

Accurately solving time-dependent partial differential equations (PDEs) with neural networks remains challenging due to long-time error accumulation and the difficulty of enforcing general boundary conditions. We introduce TENG-BC, a high-precision neural PDE solver based on the Time-Evolving Natural Gradient, designed to perform under generic boundary constraints. At each time step, TENG-BC performs a boundary-aware optimization that jointly enforces interior dynamics and boundary conditions, accommodating Dirichlet, Neumann, Robin, and mixed types within a unified framework. This formulation admits a natural-gradient interpretation, enabling stable time evolution without delicate penalty tuning. Across benchmarks over diffusion, transport, and nonlinear PDEs with various boundary conditions, TENG-BC achieves solver-level accuracy under comparable sampling budgets, outperforming conventional solvers and physics-informed neural network (PINN) baselines.

cs.LG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2611.14125

Asymptotically good codes from Siegel modular varieties

Yota Maeda

原文摘要

We construct asymptotically good codes by evaluating Siegel modular forms at superspecial points on congruence towers of Siegel modular varieties. For every genus $g\ge2$ and every fixed weight $w\ge g+1$, the construction gives such a family in sufficiently large characteristic, over a finite field which is independent of the level of the tower. The towers can be chosen so that this field is $\mathbb{F}_{p^2}$. We give an explicit lower bound for the asymptotic rate. The lower bound for the relative minimum distance follows from the Pieri formula on the space of flags in the Hodge bundle and an intersection argument along its strata. We also prove that, for $p\ge5$ and $0\le w\le p+1$, the kernel of evaluation in every genus consists precisely of the multiples of the ordinary Hasse invariant. For $g=2,3$, the geometry of the supersingular locus gives explicit bounds for the relative minimum distance and recovery sets whose sizes are bounded independently of the level.

math.NT · cs.IT · math.AG

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2403.17755

DataCook: Exploring Anti-Adversarial Fingerprints for Healthcare Data Usage Control

Sihan Shang, Jun Yan, Danli Shi, Haizhong Zheng, Shuangjia Zheng, Wei Zhao, Jiancheng Yang

原文摘要

Purpose: Healthcare AI increasingly relies on high-quality datasets, yet medical data are scarce, costly to curate, and tightly regulated in terms of data access. More critically, once medical data are leaked or accessed without authorization, data owners have limited means to prevent such data from being used to train and deploy downstream models. To address this gap, we propose DataCook, a deployment-stage usage-control framework based on anti-adversarial fingerprints. Materials and Methods: Specifically, DataCook transforms raw samples through a private cooking procedure that embeds anti-adversarial fingerprints. Models trained on the cooked data retain strong utility when evaluated on similarly cooked inputs, while their performance on raw inputs is reduced, making downstream model utility dependent on access to the private cooking function. We evaluate DataCook on 2D, 3D, and high-resolution medical images, as well as structured healthcare tabular data, across multiple model architectures and adaptive attack settings. Results: Across the evaluated settings, DataCook generally preserves authorized model utility while reducing performance on unauthorized raw inputs. The authorization-dependent behavior is observed across multiple medical imaging settings, model architectures, and adaptive attack conditions, supporting the feasibility of the proposed deployment-stage usage-control mechanism. Conclusion: These results demonstrate the feasibility of authorization-dependent model use at deployment, while the effectiveness of the current framework remains dependent on the specific task and evaluation metric.

cs.AI · cs.CR · cs.CV

查看论文
arXiv ↗PDF ↗
版本更新 · v2arXiv:2610.33070

PhaseBreak: Proactive Temporal Protection against Unauthorized Image-to-Video Animation

Jiashu Zhang, Xuanyou Li, Qixin Zhang

原文摘要

Image-to-Video (I2V) diffusion models can animate a single portrait or photograph, enabling shared images to be repurposed for motion synthesis without the subject's authorization. We study proactive visual protection against such use: before an image is released, its owner applies a bounded perturbation that preserves the sharing workflow but disrupts downstream motion generation. Existing spectral disruption objectives globally attenuate temporal representations and therefore provide limited control over subtle or spatially localized motion. Our key idea is that generated dynamics depend not only on temporal feature strength, but also on the order in which feature events occur. We propose PhaseBreak, a white-box spectral retiming protection method that matches intermediate denoiser features to a constrained temporal rearrangement of their clean counterparts. PhaseBreak operates along the true temporal axis and uses a subject-aware mask to confine the protective perturbation to the depicted subject. Evaluations on Stable Video Diffusion and DynamiCrafter show that protected images induce greater short-term motion variation and lower temporal regularity, impeding stable motion customization from the released image.

cs.CV · cs.AI

查看论文
arXiv ↗PDF ↗
版本更新 · v4arXiv:2306.04169

Efficient Alternating Minimization with Applications to Weighted Low Rank Approximation

Zhao Song, Mingquan Ye, Junze Yin, Lichen Zhang

原文摘要

Weighted low rank approximation is a fundamental problem in numerical linear algebra, and it has many applications in machine learning. Given a matrix $M \in \mathbb{R}^{n \times n}$, a non-negative weight matrix $W \in \mathbb{R}_{\geq 0}^{n \times n}$, a parameter $k$, the goal is to output two matrices $X,Y\in \mathbb{R}^{n \times k}$ such that $\| W \circ (M - X Y^\top) \|_F$ is minimized, where $\circ$ denotes the Hadamard product. It naturally generalizes the well-studied low rank matrix completion problem. Such a problem is known to be NP-hard and even hard to approximate assuming the Exponential Time Hypothesis [GG11, RSW16]. Meanwhile, alternating minimization is a good heuristic solution for weighted low rank approximation. In particular, [LLR16] shows that, under mild assumptions, alternating minimization does provide provable guarantees. In this work, we develop an efficient and robust framework for alternating minimization that allows the alternating updates to be computed approximately. For weighted low rank approximation, this improves the cost of each iteration of [LLR16] from $\|W\|_0k^2$ to $\|W\|_0 k$ where $\|W\|_0$ denotes the number of nonzero entries of the weight matrix. At the heart of our framework is a high-accuracy multiple response regression solver together with a robust analysis of alternating minimization. We also replace the random initialization of [LLR16], whose claimed guarantee fails, by subspace iteration from a Gaussian start, so our total running time of $\widetilde O((\|W\|_0 k+nk^3)\log(1/ε))$ includes the initialization. In exchange, our guarantee needs the weight matrix to be within spectral distance $γn$ of the all-ones matrix for a $γ$ that is polynomially small in $n$.

cs.LG · math.OC

查看论文
arXiv ↗PDF ↗