arXiv subjects

Yan Han

Publications and source records attributed to Yan Han.

At least 19 recordsLinked to original sources

arXiv2026

Contextual Credit Policy Optimization

Group-based reinforcement learning enables large language model agents to learn from sparse outcome rewards in long-horizon tasks. Recent methods derive step-level credit by comparing observation-matched experiences across trajectories. However, identical observations can occur in different contexts, and matched peers are not equally informative for a given decision. Moreover, favorable terminal outcomes can obscure redundant actions or detours within successful trajectories. To address these two challenges, we propose Contextual Credit Policy Optimization (CCPO). For the first, CCPO weights observation-matched peers from other trajectories by contextual similarity, computed from frozen LLM representations and accumulated trajectory statistics, allowing peers from similar contexts to contribute more to a step's baseline. For the second, CCPO pairs historical credit, which compares the realized return against the weighted peer baseline, with future credit, which measures the change in contextual outcome value over a short realized continuation. A step is thus credited for the progress it makes, not only for its trajectory's eventual outcome. Both credits are computed in hindsight from rollouts already collected for training, enabling CCPO to address these challenges without an auxiliary value critic. Across two Qwen2.5-Instruct backbones (1.5B and 7B), CCPO achieves up to 97.4% success on ALFWorld and 79.4% on WebShop and outperforms GRPO by up to 21.6 and 18.5 points. Ablations show that contextual weighting outperforms uniform weighting, and that historical and future credit serve as complementary learning signals.

cs.LG↗

arXiv2026

TouchScale: 500 Hours of Human Vision and Touch for Visual-Tactile Learning

Large-scale egocentric human interaction data is becoming an important source of physical supervision for embodied learning, yet video alone leaves the contact and pressure that characterize physical interaction unrecorded. Recent visual-tactile datasets provide this missing supervision, but their synchronized tactile data remain far smaller in volume than human video. Moreover, the largest resources often merge recordings from different sensors or annotation procedures, which makes the effect of data scale difficult to isolate. We therefore introduce TouchScale, a 500-hour dataset of contact-rich human interaction recorded with a single unified wearable setup. Its approximately 2K predefined task descriptions span everyday activities and structured manipulation, and each recording temporally aligns egocentric RGB-D video with wrist RGB video and dense full-hand bimanual tactile measurements. Compared with prior tactile data, training on the full TouchScale raises zero-shot contact IoU on data from an unseen tactile sensor from 0.134 to 0.383. Pretraining a visual encoder on TouchScale also yields the highest action recognition accuracy on three benchmarks among the compared visual-tactile datasets. Used for visual-tactile mid-training of a robot policy, TouchScale improves the average real-world success rate across four contact-rich manipulation tasks from 22.5% to 57.5%. With the sensor and collection protocol held fixed, both zero-shot tactile prediction and robot success show an overall upward trend as more TouchScale data is used. These results suggest that human visual-tactile data collected at scale with consistent sensing benefits both perception and robot manipulation. We will publicly release TouchScale, including all synchronized visual-tactile recordings and reconstructed object models, to support future research on scalable visual-tactile learning.

cs.CV↗

arXiv2026

Nonlinear pole topology of slow currents in holographic probe-brane metals

Low-frequency Drude response does not by itself determine which nearly conserved sector carries an electric current. We classify the leading cubic retarded response of an isolated, inversion-symmetric slow vector by its pole topology. Constitutive nonlinearity produces only incoming-frequency poles, whereas nonlinear relaxation adds a propagator at the emitted frequency; coupling to a second scalar slow mode generates pair-frequency poles. We then compute the full complex third-harmonic response of finite-density Dirac--Born--Infeld probe branes in Lifshitz black-brane backgrounds. In the marginal $z=2$ theory, the dense infrared response approaches the relaxation topology, with a signed infrared weight $p_{\rm IR}=0.994\pm0.002$ and a normalization-independent zero at $ω_\times/Γ_J=0.2918$. Nonmarginal $z=1$ and $z=3/2$ backgrounds instead realize mixed constitutive--relaxation weights, demonstrating that the topology is selected dynamically rather than imposed by the DBI square root. The exact nonlinear dc solution further yields a parameter-free dense-limit scaling function and a crossover field $E_{\rm nl}\propto T^{3/2}$ at $z=2$. These results connect emergent higher-form slow currents, nonlinear response theory, and holographic transport, while the two-mode extension identifies the additional pole structures expected when energy or deformation modes remain slow.

hep-th↗

arXiv2026

Beyond Artifacts: Towards Generalizable Synthetic Song Detection via Music-Intrinsic Features

The rapid advancement of AI music generators highlights the urgent need for reliable Synthetic Song Detection (SSD). Existing SSD methods often rely on low-level artifacts or fixed feature assumptions, struggling to capture generator-agnostic cues. To address this, we propose Sofia (Synthetic-song detection framework via music features), a flexible framework that models music-intrinsic attributes via feature-specific experts and an adaptive Mixture-of-Experts (MoE) module. By configuring Sofia with representative Vocal, Audio-effect, Global structure features, and their combinations, we present their individual and complementary contributions. To comprehensively evaluate our framework, we further construct MUSIC8K, a challenging benchmark featuring lastest emerging generators and realistic audio perturbations. Experiments show that Sofia learns generator-agnostic representations from music-intrinsic features, improving the F1 score by 18.5 points over the strongest baseline on MUSIC8K-O while maintaining strong robustness.

cs.SD↗

arXiv2026

WEBSERV: A Full-Stack and RL-Ready Web Environment for Training Web Agents at Scale

Reinforcement learning (RL) for web agents demands environments that are both effective for evaluation and efficient enough for large-scale on-policy training. Current web environments fall short: server-side Docker setups are too resource-intensive for massive parallel rollouts, while browser-side interfaces produce noisy observations, execute actions unreliably under modern single-page applications, and omit visual interactivity cues. We introduce WebServ, a full-stack, RL-ready web environment that addresses these limitations end-to-end. On the server side, WebServ uses Incus containers with block-level copy-on-write, reducing launch latency by ~5x and persistent storage by ~240x, enabling 200+ concurrent isolated environments on a single host. On the browser side, WebServ provides a compact, site-agnostic observation and action interface derived automatically from the DOM with human-aligned interactivity cues, and a robust action execution backend using network-aware waiting for reliable SPA support. On WebArena-Lite, WebServ achieves state-of-the-art single-prompt results, with controlled comparisons confirming consistent gains across GPT-4o, OpenAI-o3, and Llama-3.1-8B over vanilla WebArena. We further train Qwen3-4B and Qwen3-30B-A3B with RL entirely within WebServ; the RL-trained 4B model achieves 55.5% mean accuracy, surpassing both Claude 4.5 Sonnet (50.0%) and the RL-trained 8B model from WebAgent-R1 (51.8%).

cs.LG↗

arXiv2026

Restoration-Oriented Video Frame Interpolation with Region-Distinguishable Priors from SAM

In existing restoration-oriented Video Frame Interpolation (VFI) approaches, the motion estimation between neighboring frames plays a crucial role. However, the estimation accuracy in existing methods remains a challenge, primarily due to the inherent ambiguity in identifying corresponding areas in adjacent frames for interpolation. Therefore, enhancing accuracy by distinguishing different regions before motion estimation is of utmost importance. In this paper, we introduce a novel solution involving the utilization of open-world segmentation models, e.g., SAM2 (Segment Anything Model2) for frames, to derive Region-Distinguishable Priors (RDPs) in different frames. These RDPs are represented as spatial-varying Gaussian mixtures, distinguishing an arbitrary number of areas with a unified modality. RDPs can be integrated into existing motion-based VFI methods to enhance features for motion estimation, facilitated by our designed play-and-plug Hierarchical Region-aware Feature Fusion Module (HRFFM). HRFFM incorporates RDP into various hierarchical stages of VFI's encoder, using RDP-guided Feature Normalization (RDPFN) in a residual learning manner. With HRFFM and RDP, the features within VFI's encoder exhibit similar representations for matched regions in neighboring frames, thus improving the synthesis of intermediate frames. Extensive experiments demonstrate that HRFFM consistently enhances VFI performance across various scenes.

cs.CV↗

arXiv2026

Can LLM Agents Simulate Multi-Turn Human Behavior? Evidence from Real Online Customer Behavior Data

Recent research shows that LLM Agents can generate ``believable'' human behaviors via prompt-only methods, and such agents have been increasingly adopted in downstream applications. However, existing evaluation of these agents only focuses on qualitative believability (whether human raters think they are accurate), leaving open questions of whether LLM agents can accurately generate step-by-step actions mimicking a particular human's behavior in a multi-turn interaction task. In this work, we take shopping as a case study and present the first large-scale quantitative evaluation of state-of-the-art LLMs' ability to accurately simulate human behavior. Using real-world data from 31,865 online shopping sessions containing 230,965 user actions, our evaluation reveals that prompt-based LLMs (DeepSeek-R1, Llama, Claude) achieve only 11.86% accuracy in generating human actions, highlighting a substantial gap in actual behavioral accuracy. Through experiments, we also showcase that strategies as simple as fine-tuning LLMs on real human click-through data augmented with synthesized reasoning traces can greatly enhance models' performance. The fine-tuned Qwen2.5-7B achieves 17.26% action generation accuracy and 33.86% F1 score on final purchase prediction, representing substantial improvements of 5.4% and 13.85% over prompt-only baselines. This work establishes the first rigorous benchmark for human behavior simulation and provides actionable insights for developing more accurate LLM agents for future downstream applications.

cs.CL↗

arXiv2026

Rethinking the Value of Multi-Agent Workflow: A Strong Single Agent Baseline

Recent advances in LLM-based multi-agent systems (MAS) show that workflows composed of multiple LLM agents with distinct roles, tools, and communication patterns can outperform single-LLM baselines on complex tasks. However, most frameworks are homogeneous, where all agents share the same base LLM and differ only in prompts, tools, and positions in the workflow. This raises the question of whether such workflows can be simulated by a single agent through multi-turn conversations. We investigate this across seven benchmarks spanning coding, mathematics, general question answering, domain-specific reasoning, and real-world planning and tool use. Our results show that a single agent can reach the performance of homogeneous workflows with an efficiency advantage from KV cache reuse, and can even match the performance of an automatically optimized heterogeneous workflow. Building on this finding, we propose \textbf{OneFlow}, an algorithm that automatically tailors workflows for single-agent execution, reducing inference costs compared to existing automatic multi-agent design frameworks without trading off accuracy. These results position the single-LLM implementation of multi-agent workflows as a strong baseline for MAS research. We also note that single-LLM methods cannot capture heterogeneous workflows due to the lack of KV cache sharing across different LLMs, highlighting future opportunities in developing \textit{truly} heterogeneous multi-agent systems.

cs.MA↗

arXiv2025

LLaDA-Rec: Discrete Diffusion for Parallel Semantic ID Generation in Generative Recommendation

Generative recommendation represents each item as a semantic ID, i.e., a sequence of discrete tokens, and generates the next item through autoregressive decoding. While effective, existing autoregressive models face two intrinsic limitations: (1) unidirectional constraints, where causal attention restricts each token to attend only to its predecessors, hindering global semantic modeling; and (2) error accumulation, where the fixed left-to-right generation order causes prediction errors in early tokens to propagate to the predictions of subsequent token. To address these issues, we propose LLaDA-Rec, a discrete diffusion framework that reformulates recommendation as parallel semantic ID generation. By combining bidirectional attention with the adaptive generation order, the approach models inter-item and intra-item dependencies more effectively and alleviates error accumulation. Specifically, our approach comprises three key designs: (1) a parallel tokenization scheme that produces semantic IDs for bidirectional modeling, addressing the mismatch between residual quantization and bidirectional architectures; (2) two masking mechanisms at the user-history and next-item levels to capture both inter-item sequential dependencies and intra-item semantic relationships; and (3) an adapted beam search strategy for adaptive-order discrete diffusion decoding, resolving the incompatibility of standard beam search with diffusion-based generation. Experiments on three real-world datasets show that LLaDA-Rec consistently outperforms both ID-based and state-of-the-art generative recommenders, establishing discrete diffusion as a new paradigm for generative recommendation.

cs.IR↗

arXiv2025

Geometric-Aware Low-Light Image and Video Enhancement via Depth Guidance

Low-Light Enhancement (LLE) is aimed at improving the quality of photos/videos captured under low-light conditions. It is worth noting that most existing LLE methods do not take advantage of geometric modeling. We believe that incorporating geometric information can enhance LLE performance, as it provides insights into the physical structure of the scene that influences illumination conditions. To address this, we propose a Geometry-Guided Low-Light Enhancement Refine Framework (GG-LLERF) designed to assist low-light enhancement models in learning improved features for LLE by integrating geometric priors into the feature representation space. In this paper, we employ depth priors as the geometric representation. Our approach focuses on the integration of depth priors into various LLE frameworks using a unified methodology. This methodology comprises two key novel modules. First, a depth-aware feature extraction module is designed to inject depth priors into the image representation. Then, Hierarchical Depth-Guided Feature Fusion Module (HDGFFM) is formulated with a cross-domain attention mechanism, which combines depth-aware features with the original image features within the LLE model. We conducted extensive experiments on public low-light image and video enhancement benchmarks. The results illustrate that our designed framework significantly enhances existing LLE methods.

cs.CV↗

arXiv2025

The existence of non-classical orthogonal quantum Latin squares

Quantum Latin squares are a generalization of classical Latin squares in quantum field and have wide applications in unitary error bases, mutually unbiased bases, $k$-uniform states and quantum error correcting codes. In this paper, we put forward some new quantum Latin squares with special properties, such as idempotent quantum Latin square, self-orthogonal quantum Latin square, holey quantum Latin square, and the notions of orthogonality on them. We present some forceful construction methods including PBD constructions and filling in holes constructions for non-classical quantum Latin squares. As consequences, we establish the existence of non-classical 2-idempotent MOQLS$(v)$, non-classical 2, 3-MOQLS$(v)$ and non-classical SOQLS$(v)$ except possibly for several definite values.

quant-ph↗

arXiv2025

Learning from Noisy Labels with Contrastive Co-Transformer

Deep learning with noisy labels is an interesting challenge in weakly supervised learning. Despite their significant learning capacity, CNNs have a tendency to overfit in the presence of samples with noisy labels. Alleviating this issue, the well known Co-Training framework is used as a fundamental basis for our work. In this paper, we introduce a Contrastive Co-Transformer framework, which is simple and fast, yet able to improve the performance by a large margin compared to the state-of-the-art approaches. We argue the robustness of transformers when dealing with label noise. Our Contrastive Co-Transformer approach is able to utilize all samples in the dataset, irrespective of whether they are clean or noisy. Transformers are trained by a combination of contrastive loss and classification loss. Extensive experimental results on corrupted data from six standard benchmark datasets including Clothing1M, demonstrate that our Contrastive Co-Transformer is superior to existing state-of-the-art methods.

cs.CV↗

arXiv2025

Stepwise Perplexity-Guided Refinement for Efficient Chain-of-Thought Reasoning in Large Language Models

Chain-of-Thought (CoT) reasoning, which breaks down complex tasks into intermediate reasoning steps, has significantly enhanced the performance of large language models (LLMs) on challenging tasks. However, the detailed reasoning process in CoT often incurs long generation times and high computational costs, partly due to the inclusion of unnecessary steps. To address this, we propose a method to identify critical reasoning steps using perplexity as a measure of their importance: a step is deemed critical if its removal causes a significant increase in perplexity. Our method enables models to focus solely on generating these critical steps. This can be achieved through two approaches: refining demonstration examples in few-shot CoT or fine-tuning the model using selected examples that include only critical steps. Comprehensive experiments validate the effectiveness of our method, which achieves a better balance between the reasoning accuracy and efficiency of CoT.

cs.CL↗

arXiv2025

Extracting and Understanding the Superficial Knowledge in Alignment

Alignment of large language models (LLMs) with human values and preferences, often achieved through fine-tuning based on human feedback, is essential for ensuring safe and responsible AI behaviors. However, the process typically requires substantial data and computation resources. Recent studies have revealed that alignment might be attainable at lower costs through simpler methods, such as in-context learning. This leads to the question: Is alignment predominantly superficial? In this paper, we delve into this question and provide a quantitative analysis. We formalize the concept of superficial knowledge, defining it as knowledge that can be acquired through easily token restyling, without affecting the model's ability to capture underlying causal relationships between tokens. We propose a method to extract and isolate superficial knowledge from aligned models, focusing on the shallow modifications to the final token selection process. By comparing models augmented only with superficial knowledge to fully aligned models, we quantify the superficial portion of alignment. Our findings reveal that while superficial knowledge constitutes a significant portion of alignment, particularly in safety and detoxification tasks, it is not the whole story. Tasks requiring reasoning and contextual understanding still rely on deeper knowledge. Additionally, we demonstrate two practical advantages of isolated superficial knowledge: (1) it can be transferred between models, enabling efficient offsite alignment of larger models using extracted superficial knowledge from smaller models, and (2) it is recoverable, allowing for the restoration of alignment in compromised models without sacrificing performance.

cs.CL↗

arXiv2024

A Survey of Calibration Process for Black-Box LLMs

Large Language Models (LLMs) demonstrate remarkable performance in semantic understanding and generation, yet accurately assessing their output reliability remains a significant challenge. While numerous studies have explored calibration techniques, they primarily focus on White-Box LLMs with accessible parameters. Black-Box LLMs, despite their superior performance, pose heightened requirements for calibration techniques due to their API-only interaction constraints. Although recent researches have achieved breakthroughs in black-box LLMs calibration, a systematic survey of these methodologies is still lacking. To bridge this gap, we presents the first comprehensive survey on calibration techniques for black-box LLMs. We first define the Calibration Process of LLMs as comprising two interrelated key steps: Confidence Estimation and Calibration. Second, we conduct a systematic review of applicable methods within black-box settings, and provide insights on the unique challenges and connections in implementing these key steps. Furthermore, we explore typical applications of Calibration Process in black-box LLMs and outline promising future research directions, providing new perspectives for enhancing reliability and human-machine alignment. This is our GitHub link: https://github.com/LiangruXie/Calibration-Process-in-Black-Box-LLMs

cs.AI↗

arXiv2024

Towards Patronizing and Condescending Language in Chinese Videos: A Multimodal Dataset and Detector

Patronizing and Condescending Language (PCL) is a form of discriminatory toxic speech targeting vulnerable groups, threatening both online and offline safety. While toxic speech research has mainly focused on overt toxicity, such as hate speech, microaggressions in the form of PCL remain underexplored. Additionally, dominant groups' discriminatory facial expressions and attitudes toward vulnerable communities can be more impactful than verbal cues, yet these frame features are often overlooked. In this paper, we introduce the PCLMM dataset, the first Chinese multimodal dataset for PCL, consisting of 715 annotated videos from Bilibili, with high-quality PCL facial frame spans. We also propose the MultiPCL detector, featuring a facial expression detection module for PCL recognition, demonstrating the effectiveness of modality complementarity in this challenging task. Our work makes an important contribution to advancing microaggression detection within the domain of toxic speech.

cs.CV↗

arXiv2024

Distinguish Any Fake Videos: Unleashing the Power of Large-scale Data and Motion Features

The development of AI-Generated Content (AIGC) has empowered the creation of remarkably realistic AI-generated videos, such as those involving Sora. However, the widespread adoption of these models raises concerns regarding potential misuse, including face video scams and copyright disputes. Addressing these concerns requires the development of robust tools capable of accurately determining video authenticity. The main challenges lie in the dataset and neural classifier for training. Current datasets lack a varied and comprehensive repository of real and generated content for effective discrimination. In this paper, we first introduce an extensive video dataset designed specifically for AI-Generated Video Detection (GenVidDet). It includes over 2.66 M instances of both real and generated videos, varying in categories, frames per second, resolutions, and lengths. The comprehensiveness of GenVidDet enables the training of a generalizable video detector. We also present the Dual-Branch 3D Transformer (DuB3D), an innovative and effective method for distinguishing between real and generated videos, enhanced by incorporating motion information alongside visual appearance. DuB3D utilizes a dual-branch architecture that adaptively leverages and fuses raw spatio-temporal data and optical flow. We systematically explore the critical factors affecting detection performance, achieving the optimal configuration for DuB3D. Trained on GenVidDet, DuB3D can distinguish between real and generated video content with 96.77% accuracy, and strong generalization capability even for unseen types.

cs.CV↗

arXiv2024

Large-scale data extraction from the UNOS organ donor documents

In this paper we focus on three major task: 1) discussing our methods: Our method captures a portion of the data in DCD flowsheets, kidney perfusion data, and Flowsheet data captured peri-organ recovery surgery. 2) demonstrating the result: We built a comprehensive, analyzable database from 2022 OPTN data. This dataset is by far larger than any previously available even in this preliminary phase; and 3) proving that our methods can be extended to all the past OPTN data and future data. The scope of our study is all Organ Procurement and Transplantation Network (OPTN) data of the USA organ donors since 2008. The data was not analyzable in a large scale in the past because it was captured in PDF documents known as ``Attachments'', whereby every donor's information was recorded into dozens of PDF documents in heterogeneous formats. To make the data analyzable, one needs to convert the content inside these PDFs to an analyzable data format, such as a standard SQL database. In this paper we will focus on 2022 OPTN data, which consists of $\approx 400,000$ PDF documents spanning millions of pages. The entire OPTN data covers 15 years (2008--20022). This paper assumes that readers are familiar with the content of the OPTN data.

cs.CV↗