arXiv subjects

Hao Peng

Publications and source records attributed to Hao Peng.

At least 145 records · Page 8Linked to original sources

arXiv2025

Stimulated Brillouin Amplification with Flying Focus

Material damage thresholds pose a fundamental limit to chirped pulse amplification (CPA) in high-power laser systems. Plasma-based amplification via stimulated Brillouin scattering (SBS) offers a damage-free alternative, yet its effectiveness has been hindered by instabilities that constrain interaction length. In this study, we report the first experimental demonstration of SBS amplification driven by a flying focus in a 3-mm plasma channel. The flying focus is generated using chromatic aberration from spherical lenses, with its velocity precisely measured by an interferometric ionization method achieving 6.6 fs timing resolution. At a focus velocity near -c, SBS amplification is realized at pump and seed intensities more than two orders of magnitude lower than in conventional setups, yielding a conversion efficiency of 14.5%. These results validate flying focus as a powerful tool for extending interaction lengths and enabling efficient plasma-based laser amplification at reduced intensities.

physics.plasm-ph↗

arXiv2025

Low-Light Video Enhancement via Spatial-Temporal Consistent Decomposition

Low-Light Video Enhancement (LLVE) seeks to restore dynamic or static scenes plagued by severe invisibility and noise. In this paper, we present an innovative video decomposition strategy that incorporates view-independent and view-dependent components to enhance the performance of LLVE. We leverage dynamic cross-frame correspondences for the view-independent term (which primarily captures intrinsic appearance) and impose a scene-level continuity constraint on the view-dependent term (which mainly describes the shading condition) to achieve consistent and satisfactory decomposition results. To further ensure consistent decomposition, we introduce a dual-structure enhancement network featuring a cross-frame interaction mechanism. By supervising different frames simultaneously, this network encourages them to exhibit matching decomposition features. This mechanism can seamlessly integrate with encoder-decoder single-frame networks, incurring minimal additional parameter costs. Extensive experiments are conducted on widely recognized LLVE benchmarks, covering diverse scenarios. Our framework consistently outperforms existing methods, establishing a new SOTA performance.

cs.CV↗

arXiv2025

Kwai Keye-VL Technical Report

While Multimodal Large Language Models (MLLMs) demonstrate remarkable capabilities on static images, they often fall short in comprehending dynamic, information-dense short-form videos, a dominant medium in today's digital landscape. To bridge this gap, we introduce \textbf{Kwai Keye-VL}, an 8-billion-parameter multimodal foundation model engineered for leading-edge performance in short-video understanding while maintaining robust general-purpose vision-language abilities. The development of Keye-VL rests on two core pillars: a massive, high-quality dataset exceeding 600 billion tokens with a strong emphasis on video, and an innovative training recipe. This recipe features a four-stage pre-training process for solid vision-language alignment, followed by a meticulous two-phase post-training process. The first post-training stage enhances foundational capabilities like instruction following, while the second phase focuses on stimulating advanced reasoning. In this second phase, a key innovation is our five-mode ``cold-start'' data mixture, which includes ``thinking'', ``non-thinking'', ``auto-think'', ``think with image'', and high-quality video data. This mixture teaches the model to decide when and how to reason. Subsequent reinforcement learning (RL) and alignment steps further enhance these reasoning capabilities and correct abnormal model behaviors, such as repetitive outputs. To validate our approach, we conduct extensive evaluations, showing that Keye-VL achieves state-of-the-art results on public video benchmarks and remains highly competitive on general image-based tasks (Figure 1). Furthermore, we develop and release the \textbf{KC-MMBench}, a new benchmark tailored for real-world short-video scenarios, where Keye-VL shows a significant advantage.

cs.CV↗

arXiv2025

Imitation Learning for Satellite Attitude Control under Unknown Perturbations

This paper presents a novel satellite attitude control framework that integrates Soft Actor-Critic (SAC) reinforcement learning with Generative Adversarial Imitation Learning (GAIL) to achieve robust performance under various unknown perturbations. Traditional control techniques often rely on precise system models and are sensitive to parameter uncertainties and external perturbations. To overcome these limitations, we first develop a SAC-based expert controller that demonstrates improved resilience against actuator failures, sensor noise, and attitude misalignments, outperforming our previous results in several challenging scenarios. We then use GAIL to train a learner policy that imitates the expert's trajectories, thereby reducing training costs and improving generalization through expert demonstrations. Preliminary experiments under single and combined perturbations show that the SAC expert can rotate the antenna to a specified direction and keep the antenna orientation reliably stable in most of the listed perturbations. Additionally, the GAIL learner can imitate most of the features from the trajectories generated by the SAC expert. Comparative evaluations and ablation studies confirm the effectiveness of the SAC algorithm and reward shaping. The integration of GAIL further reduces sample complexity and demonstrates promising imitation capabilities, paving the way for more intelligent and autonomous spacecraft control systems.

eess.SY↗

arXiv2025

FactCheckmate: Preemptively Detecting and Mitigating Hallucinations in LMs

Language models (LMs) hallucinate. We inquire: Can we detect and mitigate hallucinations before they happen? This work answers this research question in the positive, by showing that the internal representations of LMs provide rich signals that can be used for this purpose. We introduce FactCheckmate, which preemptively detects hallucinations by learning a classifier that predicts whether the LM will hallucinate, based on the model's hidden states produced over the inputs, before decoding begins. If a hallucination is detected, FactCheckmate then intervenes by adjusting the LM's hidden states such that the model will produce more factual outputs. FactCheckmate provides fresh insights that the inner workings of LMs can be revealed by their hidden states. Practically, both its detection and mitigation models are lightweight, adding little inference overhead; FactCheckmate proves a more efficient approach for mitigating hallucinations compared to many post-hoc alternatives. We evaluate FactCheckmate over LMs of different scales and model families (including Llama, Mistral, Qwen and Gemma), across a variety of QA datasets from different domains. Our results demonstrate the effectiveness of FactCheckmate, achieving over 70% preemptive detection accuracy. On average, outputs generated by LMs with intervention are 34.4% more factual compared to those without.

cs.CL↗

arXiv2025

Hierarchical Decision Making Based on Structural Information Principles

Hierarchical Reinforcement Learning (HRL) is a promising approach for managing task complexity across multiple levels of abstraction and accelerating long-horizon agent exploration. However, the effectiveness of hierarchical policies heavily depends on prior knowledge and manual assumptions about skill definitions and task decomposition. In this paper, we propose a novel Structural Information principles-based framework, namely SIDM, for hierarchical Decision Making in both single-agent and multi-agent scenarios. Central to our work is the utilization of structural information embedded in the decision-making process to adaptively and dynamically discover and learn hierarchical policies through environmental abstractions. Specifically, we present an abstraction mechanism that processes historical state-action trajectories to construct abstract representations of states and actions. We define and optimize directed structural entropy, a metric quantifying the uncertainty in transition dynamics between abstract states, to discover skills that capture key transition patterns in RL environments. Building on these findings, we develop a skill-based learning method for single-agent scenarios and a role-based collaboration method for multi-agent scenarios, both of which can flexibly integrate various underlying algorithms for enhanced performance. Extensive evaluations on challenging benchmarks demonstrate that our framework significantly and consistently outperforms state-of-the-art baselines, improving the effectiveness, efficiency, and stability of policy learning by up to 32.70%, 64.86%, and 88.26%, respectively, as measured by average rewards, convergence timesteps, and standard deviations.

cs.LG↗

arXiv2025

Exploring Strategies for Personalized Radiation Therapy Part I Unlocking Response-Related Tumor Subregions with Class Activation Mapping

Personalized precision radiation therapy requires more than simple classification, it demands the identification of prognostic, spatially informative features and the ability to adapt treatment based on individual response. This study compares three approaches for predicting treatment response: standard radiomics, gradient based features, and convolutional neural networks enhanced with Class Activation Mapping. We analyzed 69 brain metastases from 39 patients treated with Gamma Knife radiosurgery. An integrated autoencoder classifier model was used to predict whether tumor volume would shrink by more than 20 percent at a three months follow up, framed as a binary classification task. The results highlight their strength in hierarchical feature extraction and the classifiers discriminative capacity. Among the models, pixel wise CAM provides the most detailed spatial insight, identifying lesion specific regions rather than relying on fixed patterns, demonstrating strong generalization. In non responding lesions, the activated regions may indicate areas of radio resistance. Pixel wise CAM outperformed both radiomics and gradient based methods in classification accuracy. Moreover, its fine grained spatial features allow for alignment with cellular level data, supporting biological validation and deeper understanding of heterogeneous treatment responses. Although further validation is necessary, these findings underscore the promise in guiding personalized and adaptive radiotherapy strategies for both photon and particle therapies.

physics.med-ph↗

arXiv2025

Exploring Strategies for Personalized Radiation Therapy Part II Predicting Tumor Drift Patterns with Diffusion Models

Radiation therapy outcomes are decided by two key parameters, dose and timing, whose best values vary substantially across patients. This variability is especially critical in the treatment of brain cancer, where fractionated or staged stereotactic radiosurgery improves safety compared to single fraction approaches, but complicates the ability to predict treatment response. To address this challenge, we employ Personalized Ultra-fractionated Stereotactic Adaptive Radiotherapy (PULSAR), a strategy that dynamically adjusts treatment based on how each tumor evolves over time. However, the success of PULSAR and other adaptive approaches depends on predictive tools that can guide early treatment decisions and avoid both overtreatment and undertreatment. However, current radiomics and dosiomics models offer limited insight into the evolving spatial and temporal patterns of tumor response. To overcome these limitations, we propose a novel framework using Denoising Diffusion Implicit Models (DDIM), which learns data-driven mappings from pre to post treatment imaging. In this study, we developed single step and iterative denoising strategies and compared their performance. The results show that diffusion models can effectively simulate patient specific tumor evolution and localize regions associated with treatment response. The proposed strategy provides a promising foundation for modeling heterogeneous treatment response and enabling early, adaptive interventions, paving the way toward more personalized and biologically informed radiotherapy.

physics.med-ph↗

arXiv2025

StoryWriter: A Multi-Agent Framework for Long Story Generation

Long story generation remains a challenge for existing large language models (LLMs), primarily due to two main factors: (1) discourse coherence, which requires plot consistency, logical coherence, and completeness in the long-form generation, and (2) narrative complexity, which requires an interwoven and engaging narrative. To address these challenges, we propose StoryWriter, a multi-agent story generation framework, which consists of three main modules: (1) outline agent, which generates event-based outlines containing rich event plots, character, and event-event relationships. (2) planning agent, which further details events and plans which events should be written in each chapter to maintain an interwoven and engaging story. (3) writing agent, which dynamically compresses the story history based on the current event to generate and reflect new plots, ensuring the coherence of the generated story. We conduct both human and automated evaluation, and StoryWriter significantly outperforms existing story generation baselines in both story quality and length. Furthermore, we use StoryWriter to generate a dataset, which contains about $6,000$ high-quality long stories, with an average length of $8,000$ words. We train the model Llama3.1-8B and GLM4-9B using supervised fine-tuning on LongStory and develop StoryWriter_GLM and StoryWriter_GLM, which demonstrates advanced performance in long story generation.

cs.CL↗

arXiv2025

STAMImputer: Spatio-Temporal Attention MoE for Traffic Data Imputation

Traffic data imputation is fundamentally important to support various applications in intelligent transportation systems such as traffic flow prediction. However, existing time-to-space sequential methods often fail to effectively extract features in block-wise missing data scenarios. Meanwhile, the static graph structure for spatial feature propagation significantly constrains the models flexibility in handling the distribution shift issue for the nonstationary traffic data. To address these issues, this paper proposes a SpatioTemporal Attention Mixture of experts network named STAMImputer for traffic data imputation. Specifically, we introduce a Mixture of Experts (MoE) framework to capture latent spatio-temporal features and their influence weights, effectively imputing block missing. A novel Low-rank guided Sampling Graph ATtention (LrSGAT) mechanism is designed to dynamically balance the local and global correlations across road networks. The sampled attention vectors are utilized to generate dynamic graphs that capture real-time spatial correlations. Extensive experiments are conducted on four traffic datasets for evaluation. The result shows STAMImputer achieves significantly performance improvement compared with existing SOTA approaches. Our codes are available at https://github.com/RingBDStack/STAMImupter.

cs.LG↗

arXiv2025

VerIF: Verification Engineering for Reinforcement Learning in Instruction Following

Reinforcement learning with verifiable rewards (RLVR) has become a key technique for enhancing large language models (LLMs), with verification engineering playing a central role. However, best practices for RL in instruction following remain underexplored. In this work, we explore the verification challenge in RL for instruction following and propose VerIF, a verification method that combines rule-based code verification with LLM-based verification from a large reasoning model (e.g., QwQ-32B). To support this approach, we construct a high-quality instruction-following dataset, VerInstruct, containing approximately 22,000 instances with associated verification signals. We apply RL training with VerIF to two models, achieving significant improvements across several representative instruction-following benchmarks. The trained models reach state-of-the-art performance among models of comparable size and generalize well to unseen constraints. We further observe that their general capabilities remain unaffected, suggesting that RL with VerIF can be integrated into existing RL recipes to enhance overall model performance. We have released our datasets, codes, and models to facilitate future research at https://github.com/THU-KEG/VerIF.

cs.CL↗

arXiv2025

Decompose, Plan in Parallel, and Merge: A Novel Paradigm for Large Language Models based Planning with Multiple Constraints

Despite significant advances in Large Language Models (LLMs), planning tasks still present challenges for LLM-based agents. Existing planning methods face two key limitations: heavy constraints and cascading errors. To address these limitations, we propose a novel parallel planning paradigm, which Decomposes, Plans for subtasks in Parallel, and Merges subplans into a final plan (DPPM). Specifically, DPPM decomposes the complex task based on constraints into subtasks, generates the subplan for each subtask in parallel, and merges them into a global plan. In addition, our approach incorporates a verification and refinement module, enabling error correction and conflict resolution. Experimental results demonstrate that DPPM significantly outperforms existing methods in travel planning tasks.

cs.CL↗

arXiv2025

SEA: Low-Resource Safety Alignment for Multimodal Large Language Models via Synthetic Embeddings

Multimodal Large Language Models (MLLMs) have serious security vulnerabilities.While safety alignment using multimodal datasets consisting of text and data of additional modalities can effectively enhance MLLM's security, it is costly to construct these datasets. Existing low-resource security alignment methods, including textual alignment, have been found to struggle with the security risks posed by additional modalities. To address this, we propose Synthetic Embedding augmented safety Alignment (SEA), which optimizes embeddings of additional modality through gradient updates to expand textual datasets. This enables multimodal safety alignment training even when only textual data is available. Extensive experiments on image, video, and audio-based MLLMs demonstrate that SEA can synthesize a high-quality embedding on a single RTX3090 GPU within 24 seconds. SEA significantly improves the security of MLLMs when faced with threats from additional modalities. To assess the security risks introduced by video and audio, we also introduced a new benchmark called VA-SafetyBench. High attack success rates across multiple MLLMs validate its challenge. Our code and data will be available at https://github.com/ZeroNLP/SEA.

cs.CL↗

arXiv2025

Probing the Geometry of Truth: Consistency and Generalization of Truth Directions in LLMs Across Logical Transformations and Question Answering Tasks

Large language models (LLMs) are trained on extensive datasets that encapsulate substantial world knowledge. However, their outputs often include confidently stated inaccuracies. Earlier works suggest that LLMs encode truthfulness as a distinct linear feature, termed the "truth direction", which can classify truthfulness reliably. We address several open questions about the truth direction: (i) whether LLMs universally exhibit consistent truth directions; (ii) whether sophisticated probing techniques are necessary to identify truth directions; and (iii) how the truth direction generalizes across diverse contexts. Our findings reveal that not all LLMs exhibit consistent truth directions, with stronger representations observed in more capable models, particularly in the context of logical negation. Additionally, we demonstrate that truthfulness probes trained on declarative atomic statements can generalize effectively to logical transformations, question-answering tasks, in-context learning, and external knowledge sources. Finally, we explore the practical application of truthfulness probes in selective question-answering, illustrating their potential to improve user trust in LLM outputs. These results advance our understanding of truth directions and provide new insights into the internal representations of LLM beliefs. Our code is public at https://github.com/colored-dye/truthfulness_probe_generalization

cs.CL↗

arXiv2025

Scaling Diffusion Language Models via Adaptation from Autoregressive Models

Diffusion Language Models (DLMs) have emerged as a promising new paradigm for text generative modeling, potentially addressing limitations of autoregressive (AR) models. However, current DLMs have been studied at a smaller scale compared to their AR counterparts and lack fair comparison on language modeling benchmarks. Additionally, training diffusion models from scratch at scale remains challenging. Given the prevalence of open-source AR language models, we propose adapting these models to build text diffusion models. We demonstrate connections between AR and diffusion modeling objectives and introduce a simple continual pre-training approach for training diffusion models. Through systematic evaluation on language modeling, reasoning, and commonsense benchmarks, we show that we can convert AR models ranging from 127M to 7B parameters (GPT2 and LLaMA) into diffusion models DiffuGPT and DiffuLLaMA, using less than 200B tokens for training. Our experimental results reveal that these models outperform earlier DLMs and are competitive with their AR counterparts. We release a suite of DLMs (127M-355M-7B) capable of generating fluent text, performing in-context learning, filling in the middle without prompt re-ordering, and following instructions https://github.com/HKUNLP/DiffuLLaMA.

cs.CL↗

arXiv2025

The Entropy Mechanism of Reinforcement Learning for Reasoning Language Models

This paper aims to overcome a major obstacle in scaling RL for reasoning with LLMs, namely the collapse of policy entropy. Such phenomenon is consistently observed across vast RL runs without entropy intervention, where the policy entropy dropped sharply at the early training stage, this diminished exploratory ability is always accompanied with the saturation of policy performance. In practice, we establish a transformation equation R=-a*e^H+b between entropy H and downstream performance R. This empirical law strongly indicates that, the policy performance is traded from policy entropy, thus bottlenecked by its exhaustion, and the ceiling is fully predictable H=0, R=-a+b. Our finding necessitates entropy management for continuous exploration toward scaling compute for RL. To this end, we investigate entropy dynamics both theoretically and empirically. Our derivation highlights that, the change in policy entropy is driven by the covariance between action probability and the change in logits, which is proportional to its advantage when using Policy Gradient-like algorithms. Empirical study shows that, the values of covariance term and entropy differences matched exactly, supporting the theoretical conclusion. Moreover, the covariance term stays mostly positive throughout training, further explaining why policy entropy would decrease monotonically. Through understanding the mechanism behind entropy dynamics, we motivate to control entropy by restricting the update of high-covariance tokens. Specifically, we propose two simple yet effective techniques, namely Clip-Cov and KL-Cov, which clip and apply KL penalty to tokens with high covariances respectively. Experiments show that these methods encourage exploration, thus helping policy escape entropy collapse and achieve better downstream performance.

cs.LG↗

arXiv2025

RAG-RL: Advancing Retrieval-Augmented Generation via RL and Curriculum Learning

Retrieval-augmented generation (RAG) systems rely on retrieval models for identifying relevant contexts and answer generation models for utilizing those contexts. However, retrievers exhibit imperfect recall and precision, limiting downstream performance. We introduce RAG-RL, an answer generation model trained not only to produce answers but also to identify and cite relevant information from larger sets of retrieved contexts, shifting some of the burden of identifying relevant documents from the retriever to the answer generator. Our approach uses curriculum learning, where the model is first trained on easier examples that include only relevant contexts. Our experiments show that these training samples enable models to acquire citation and reasoning skills with greater sample efficiency and generalizability, demonstrating strong model performance even as the number of irrelevant passages increases. We benchmark our methods on three open-domain multi-hop question answering datasets and report significant gains in answer and citation accuracy. Our experiments provide empirical insights into how easier training samples can give models stronger signals for learning specific skills (e.g., citation generation) and how different components of post-training (e.g., training set construction, rule-based rewards, training sample ordering, etc.) impact final model performance.

cs.CL↗

arXiv2025

AGENTIF: Benchmarking Instruction Following of Large Language Models in Agentic Scenarios

Large Language Models (LLMs) have demonstrated advanced capabilities in real-world agentic applications. Growing research efforts aim to develop LLM-based agents to address practical demands, introducing a new challenge: agentic scenarios often involve lengthy instructions with complex constraints, such as extended system prompts and detailed tool specifications. While adherence to such instructions is crucial for agentic applications, whether LLMs can reliably follow them remains underexplored. In this paper, we introduce AgentIF, the first benchmark for systematically evaluating LLM instruction following ability in agentic scenarios. AgentIF features three key characteristics: (1) Realistic, constructed from 50 real-world agentic applications. (2) Long, averaging 1,723 words with a maximum of 15,630 words. (3) Complex, averaging 11.9 constraints per instruction, covering diverse constraint types, such as tool specifications and condition constraints. To construct AgentIF, we collect 707 human-annotated instructions across 50 agentic tasks from industrial application agents and open-source agentic systems. For each instruction, we annotate the associated constraints and corresponding evaluation metrics, including code-based evaluation, LLM-based evaluation, and hybrid code-LLM evaluation. We use AgentIF to systematically evaluate existing advanced LLMs. We observe that current models generally perform poorly, especially in handling complex constraint structures and tool specifications. We further conduct error analysis and analytical experiments on instruction length and meta constraints, providing some findings about the failure modes of existing LLMs. We have released the code and data to facilitate future research.

cs.AI↗