arXiv subjects

Xuanjing Huang

Publications and source records attributed to Xuanjing Huang.

At least 109 records · Page 6Linked to original sources

arXiv2026

Benchmark^2: Systematic Evaluation of LLM Benchmarks

The rapid proliferation of benchmarks for evaluating large language models (LLMs) has created an urgent need for systematic methods to assess benchmark quality itself. We propose Benchmark^2, a comprehensive framework comprising three complementary metrics: (1) Cross-Benchmark Ranking Consistency, measuring whether a benchmark produces model rankings aligned with peer benchmarks; (2) Discriminability Score, quantifying a benchmark's ability to differentiate between models; and (3) Capability Alignment Deviation, identifying problematic instances where stronger models fail but weaker models succeed within the same model family. We conduct extensive experiments across 15 benchmarks spanning mathematics, reasoning, and knowledge domains, evaluating 11 LLMs across four model families. Our analysis reveals significant quality variations among existing benchmarks and demonstrates that selective benchmark construction based on our metrics can achieve comparable evaluation performance with substantially reduced test sets.

cs.CL↗

arXiv2026

CMDAR: A Chinese Multi-scene Dynamic Audio Reasoning Benchmark with Diverse Challenges

The ability to reason from audio, including speech, environmental sounds, and music, is essential for AI agents to interact effectively in real-world scenarios. Existing benchmarks mainly focus on static or single-scene settings and English audio data and do not fully capture scenarios where multiple speakers, unfolding events, and heterogeneous audio sources interact. To address these challenges, we introduce CMDAR, a Chinese benchmark for evaluating models on complex, multi-scene, and dynamically evolving audio reasoning tasks. CMDAR comprises 3,000 carefully curated question-answer pairs linked to diverse audio clips, covering five categories of complex reasoning and spanning three question types. We benchmark 26 state-of-the-art audio language models on CMDAR and observe that they exhibit limitations in complex reasoning tasks. In CMDAR-main, Qwen2.5-Omni achieves 76.67% accuracy, whereas GPT-4o Audio reaches 68.47%. However, GPT-4o Audio substantially outperforms Qwen2.5-Omni on the more challenging multiple-choice with multiple audios and open-ended tasks. And we provide detail analysis corresponding suggestions for the future development of large audio language models.

cs.SD↗

arXiv2026

Multi-hop Reasoning via Early Knowledge Alignment

Retrieval-Augmented Generation (RAG) has emerged as a powerful paradigm for Large Language Models (LLMs) to address knowledge-intensive queries requiring domain-specific or up-to-date information. To handle complex multi-hop questions that are challenging for single-step retrieval, iterative RAG approaches incorporating reinforcement learning have been proposed. However, existing iterative RAG systems typically plan to decompose questions without leveraging information about the available retrieval corpus, leading to inefficient retrieval and reasoning chains that cascade into suboptimal performance. In this paper, we introduce Early Knowledge Alignment (EKA), a simple but effective module that aligns LLMs with retrieval set before planning in iterative RAG systems with contextually relevant retrieved knowledge. Extensive experiments on six standard RAG datasets demonstrate that by establishing a stronger reasoning foundation, EKA significantly improves retrieval precision, reduces cascading errors, and enhances both performance and efficiency. Our analysis from an entropy perspective demonstrate that incorporating early knowledge reduces unnecessary exploration during the reasoning process, enabling the model to focus more effectively on relevant information subsets. Moreover, EKA proves effective as a versatile, training-free inference strategy that scales seamlessly to large models. Generalization tests across diverse datasets and retrieval corpora confirm the robustness of our approach. Overall, EKA advances the state-of-the-art in iterative RAG systems while illuminating the critical interplay between structured reasoning and efficient exploration in reinforcement learning-augmented frameworks. The code is released at \href{https://github.com/yxzwang/EarlyKnowledgeAlignment}{Github}.

cs.CL↗

arXiv2025

Game-RL: Synthesizing Multimodal Verifiable Game Data to Boost VLMs' General Reasoning

Vision-language reinforcement learning (RL) has primarily focused on narrow domains (e.g. geometry or chart reasoning). This leaves broader training scenarios and resources underexplored, limiting the exploration and learning of Vision Language Models (VLMs) through RL. We find video games inherently provide rich visual elements and mechanics that are easy to verify. To fully use the multimodal and verifiable reward in video games, we propose Game-RL, constructing diverse game tasks for RL training to boost VLMs general reasoning ability. To obtain training data, we propose Code2Logic, a novel approach that adapts game code to synthesize game reasoning task data, thus obtaining the GameQA dataset of 30 games and 158 tasks with controllable difficulty gradation. Unexpectedly, RL training solely on GameQA enables multiple VLMs to achieve performance improvements across 7 diverse vision-language benchmarks, demonstrating the value of Game-RL for enhancing VLMs' general reasoning. Furthermore, this suggests that video games may serve as valuable scenarios and resources to boost general reasoning abilities. Our code, dataset and models are available at the GitHub repository.

cs.CL↗

arXiv2025

Domain-RAG: Retrieval-Guided Compositional Image Generation for Cross-Domain Few-Shot Object Detection

Cross-Domain Few-Shot Object Detection (CD-FSOD) aims to detect novel objects with only a handful of labeled samples from previously unseen domains. While data augmentation and generative methods have shown promise in few-shot learning, their effectiveness for CD-FSOD remains unclear due to the need for both visual realism and domain alignment. Existing strategies, such as copy-paste augmentation and text-to-image generation, often fail to preserve the correct object category or produce backgrounds coherent with the target domain, making them non-trivial to apply directly to CD-FSOD. To address these challenges, we propose Domain-RAG, a training-free, retrieval-guided compositional image generation framework tailored for CD-FSOD. Domain-RAG consists of three stages: domain-aware background retrieval, domain-guided background generation, and foreground-background composition. Specifically, the input image is first decomposed into foreground and background regions. We then retrieve semantically and stylistically similar images to guide a generative model in synthesizing a new background, conditioned on both the original and retrieved contexts. Finally, the preserved foreground is composed with the newly generated domain-aligned background to form the generated image. Without requiring any additional supervision or training, Domain-RAG produces high-quality, domain-consistent samples across diverse tasks, including CD-FSOD, remote sensing FSOD, and camouflaged FSOD. Extensive experiments show consistent improvements over strong baselines and establish new state-of-the-art results. Codes will be released upon acceptance.

cs.CV↗

arXiv2025

The Role of Entropy in Visual Grounding: Analysis and Optimization

Recent advances in fine-tuning multimodal large language models (MLLMs) using reinforcement learning have achieved remarkable progress, particularly with the introduction of various entropy control techniques. However, the role and characteristics of entropy in perception-oriented tasks like visual grounding, as well as effective strategies for controlling it, remain largely unexplored. To address this issue, we focus on the visual grounding task and analyze the role and characteristics of entropy in comparison to reasoning tasks. Building on these findings, we introduce ECVGPO (Entropy Control Visual Grounding Policy Optimization), an interpretable algorithm designed for effective entropy regulation. Through entropy control, the trade-off between exploration and exploitation is better balanced. Experiments show that ECVGPO achieves broad improvements across various benchmarks and models.

cs.CV↗

arXiv2025

Nex-N1: Agentic Models Trained via a Unified Ecosystem for Large-Scale Environment Construction

The evolution of Large Language Models (LLMs) from passive responders to autonomous agents necessitates a fundamental shift in learning paradigms -- from static imitation to incentive-driven decision making. However, this transition is significantly impeded by the lack of scalable infrastructure capable of constructing high-quality interaction signals for effective policy learning. To address this, we introduce a comprehensive method designed to systematically scale the diversity and complexity of interactive environments. Our method realizes this scaling by addressing three orthogonal dimensions: (1) Complexity: NexAU, a flexible agent framework that supports building complex agent hierarchies via simple configurations; (2) Diversity: NexA4A automatically generates diverse agent hierarchies from natural language to cover infinite domains; and (3) Fidelity: NexGAP bridges the simulation-reality gap by integrating dynamic real-world environment for grounded trajectories synthesis. We train Nex-N1 upon the diverse and complex interactive environments established by our infrastructure. Empirical results on benchmarks such as SWE-bench and tau2 demonstrate that Nex-N1 consistently outperforms SOTA open-source models and achieves competitive performance against frontier proprietary models on complex agentic tasks. We open-source the Nex ecosystem and model weights to facilitate further research.

cs.CL↗

arXiv2025

Thus Spake Long-Context Large Language Model

Long context is an important topic in Natural Language Processing (NLP), running through the development of NLP architectures, and offers immense opportunities for Large Language Models (LLMs), giving LLMs the lifelong learning potential akin to humans. Unfortunately, the pursuit of a long context is accompanied by numerous obstacles. Nevertheless, long context remains a core competitive advantage for LLMs. In the past two years, the context length of LLMs has achieved a breakthrough extension to millions of tokens. Moreover, research on long-context LLMs has expanded beyond length extrapolation to a comprehensive focus on architecture, infrastructure, training, and evaluation technologies. Inspired by the symphonic poem, Thus Spake Zarathustra, we draw an analogy between the journey of extending the context of LLM and the attempts of humans to transcend their mortality. In this survey, we will illustrate how LLM struggles between the tremendous need for a longer context and its equal need to accept the fact that it is ultimately finite. To achieve this, we give a global picture of the lifecycle of long-context LLMs from four perspectives: architecture, infrastructure, training, and evaluation, showcasing the full spectrum of long-context technologies. At the end of this survey, we will present 10 unanswered questions currently faced by long-context LLMs. We hope this survey can serve as a systematic introduction to research on long-context LLMs. Video: https://www.bilibili.com/video/BV11h9AYoEYj. Github: https://github.com/OpenMOSS/Thus-Spake-Long-Context-LLM.

cs.CL↗

arXiv2025

AgentPRM: Process Reward Models for LLM Agents via Step-Wise Promise and Progress

Despite rapid development, large language models (LLMs) still encounter challenges in multi-turn decision-making tasks (i.e., agent tasks) like web shopping and browser navigation, which require making a sequence of intelligent decisions based on environmental feedback. Previous work for LLM agents typically relies on elaborate prompt engineering or fine-tuning with expert trajectories to improve performance. In this work, we take a different perspective: we explore constructing process reward models (PRMs) to evaluate each decision and guide the agent's decision-making process. Unlike LLM reasoning, where each step is scored based on correctness, actions in agent tasks do not have a clear-cut correctness. Instead, they should be evaluated based on their proximity to the goal and the progress they have made. Building on this insight, we propose a re-defined PRM for agent tasks, named AgentPRM, to capture both the interdependence between sequential decisions and their contribution to the final goal. This enables better progress tracking and exploration-exploitation balance. To scalably obtain labeled data for training AgentPRM, we employ a Temporal Difference-based (TD-based) estimation method combined with Generalized Advantage Estimation (GAE), which proves more sample-efficient than prior methods. Extensive experiments across different agentic tasks show that AgentPRM is over $8\times$ more compute-efficient than baselines, and it demonstrates robust improvement when scaling up test-time compute. Moreover, we perform detailed analyses to show how our method works and offer more insights, e.g., applying AgentPRM to the reinforcement learning of LLM agents.

cs.CL↗

arXiv2025

Revealing emergent human-like conceptual representations from language prediction

People acquire concepts through rich physical and social experiences and use them to understand and navigate the world. In contrast, large language models (LLMs), trained solely through next-token prediction on text, exhibit strikingly human-like behaviors. Are these models developing concepts akin to those of humans? If so, how are such concepts represented, organized, and related to behavior? Here, we address these questions by investigating the representations formed by LLMs during an in-context concept inference task. We found that LLMs can flexibly derive concepts from linguistic descriptions in relation to contextual cues about other concepts. The derived representations converge toward a shared, context-independent structure, and alignment with this structure reliably predicts model performance across various understanding and reasoning tasks. Moreover, the convergent representations effectively capture human behavioral judgments and closely align with neural activity patterns in the human brain, providing evidence for biological plausibility. Together, these findings establish that structured, human-like conceptual representations can emerge purely from language prediction without real-world grounding, highlighting the role of conceptual structure in understanding intelligent behavior. More broadly, our work suggests that LLMs offer a tangible window into the nature of human concepts and lays the groundwork for advancing alignment between artificial and human intelligence.

cs.CL↗

arXiv2025

Compression Hacking: A Supplementary Perspective on Informatics Properties of Language Models from Geometric Distortion

Recently, the concept of ``compression as intelligence'' has provided a novel informatics metric perspective for language models (LMs), emphasizing that highly structured representations signify the intelligence level of LMs. However, from a geometric standpoint, the word representation space of highly compressed LMs tends to degenerate into a highly anisotropic state, which hinders the LM's ability to comprehend instructions and directly impacts its performance. We found this compression-anisotropy synchronicity is essentially the ``Compression Hacking'' in LM representations, where noise-dominated directions tend to create the illusion of high compression rates by sacrificing spatial uniformity. Based on this, we propose three refined compression metrics by incorporating geometric distortion analysis and integrate them into a self-evaluation pipeline. The refined metrics exhibit strong alignment with the LM's comprehensive capabilities, achieving Spearman correlation coefficients above 0.9, significantly outperforming both the original compression and other internal structure-based metrics. This confirms that compression hacking substantially enhances the informatics interpretation of LMs by incorporating geometric distortion of representations.

cs.CL↗

arXiv2025

Counteracting Matthew Effect in Self-Improvement of LVLMs through Head-Tail Re-balancing

Self-improvement has emerged as a mainstream paradigm for advancing the reasoning capabilities of large vision-language models (LVLMs), where models explore and learn from successful trajectories iteratively. However, we identify a critical issue during this process: the model excels at generating high-quality trajectories for simple queries (i.e., head data) but struggles with more complex ones (i.e., tail data). This leads to an imbalanced optimization that drives the model to prioritize simple reasoning skills, while hindering its ability to tackle more complex reasoning tasks. Over iterations, this imbalance becomes increasingly pronounced--a dynamic we term the "Matthew effect"--which ultimately hinders further model improvement and leads to performance bottlenecks. To counteract this challenge, we introduce four efficient strategies from two perspectives: distribution-reshaping and trajectory-resampling, to achieve head-tail re-balancing during the exploration-and-learning self-improvement process. Extensive experiments on Qwen2-VL-7B-Instruct and InternVL2.5-4B models across visual reasoning tasks demonstrate that our methods consistently improve visual reasoning capabilities, outperforming vanilla self-improvement by 3.86 points on average.

cs.CV↗

arXiv2025

Parrot: A Training Pipeline Enhances Both Program CoT and Natural Language CoT for Reasoning

Natural language chain-of-thought (N-CoT) and Program chain-of-thought (P-CoT) have emerged as two primary paradigms for large language models (LLMs) to solve mathematical reasoning problems. Current research typically endeavors to achieve unidirectional enhancement: P-CoT enhanced N-CoT or N-CoT enhanced P-CoT. In this paper, we seek to fully unleash the two paradigms' strengths for mutual enhancement and ultimately achieve simultaneous improvements. We conduct a detailed analysis of the error types across two paradigms, based on which we propose Parrot, a novel training pipeline for mathematical problems: 1) Three target-designed subtasks integrate sequential P-CoT and N-CoT generation. 2) A subtask hybrid training strategy to facilitate natural language semantic transferability. 3) The converted N-CoT auxiliary reward is designed to alleviate the sparse rewards in P-CoT optimization. Extensive experiments demonstrate that Parrot significantly enhances both the performance of N-CoT and P-CoT, especially on N-CoT. Using Parrot SFT, the N-CoT performance of LLaMA2 and CodeLLaMA achieve gains of +21.87 and +21.48 on MathQA over the RL baseline, which is resource-intensive.

cs.CL↗

arXiv2025

Critique-RL: Training Language Models for Critiquing through Two-Stage Reinforcement Learning

Training critiquing language models to assess and provide feedback on model outputs is a promising way to improve LLMs for complex reasoning tasks. However, existing approaches typically rely on stronger supervisors for annotating critique data. To address this, we propose Critique-RL, an online RL approach for developing critiquing language models without stronger supervision. Our approach operates on a two-player paradigm: the actor generates a response, the critic provides feedback, and the actor refines the response accordingly. We first reveal that relying solely on indirect reward signals from the actor's outputs for RL optimization often leads to unsatisfactory critics: while their helpfulness (i.e., providing constructive feedback) improves, the discriminability (i.e., determining whether a response is high-quality or not) remains poor, resulting in marginal performance gains. To overcome this, Critique-RL adopts a two-stage optimization strategy. In stage I, it reinforces the discriminability of the critic with direct rule-based reward signals; in stage II, it introduces indirect rewards based on actor refinement to improve the critic's helpfulness, while maintaining its discriminability via appropriate regularization. Extensive experiments across various tasks and models show that Critique-RL delivers substantial performance improvements. For example, it achieves a 9.02% gain on in-domain tasks and a 5.70% gain on out-of-domain tasks for Qwen2.5-7B, highlighting its potential.

cs.CL↗

arXiv2025

EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving

We introduce EvaLearn, a pioneering benchmark designed to evaluate large language models (LLMs) on their learning capability and efficiency in challenging tasks, a critical, yet underexplored aspect of model potential. EvaLearn contains 648 challenging problems across six task types, grouped into 182 sequences, each sequence dedicated to one task type. Diverging from most existing benchmarks that evaluate models in parallel, EvaLearn requires models to solve problems sequentially, allowing them to leverage the experience gained from previous solutions. EvaLearn provides five comprehensive automated metrics to evaluate models and quantify their learning capability and efficiency. We extensively benchmark nine frontier models and observe varied performance profiles: some models, such as Claude-3.7-sonnet, start with moderate initial performance but exhibit strong learning ability, while some models struggle to benefit from experience and may even show negative transfer. Moreover, we investigate model performance under two learning settings and find that instance-level rubrics and teacher-model feedback further facilitate model learning. Importantly, we observe that current LLMs with stronger static abilities do not show a clear advantage in learning capability across all tasks, highlighting that EvaLearn evaluates a new dimension of model performance. We hope EvaLearn provides a novel evaluation perspective for assessing LLM potential and understanding the gap between models and human capabilities, promoting the development of deeper and more dynamic evaluation approaches. All datasets, the automatic evaluation framework, and the results studied in this paper are available at the GitHub repository.

cs.CL↗

arXiv2025

R-Horizon: How Far Can Your Large Reasoning Model Really Go in Breadth and Depth?

Recent trends in test-time scaling for reasoning models (e.g., OpenAI o1, DeepSeek-R1) have led to remarkable improvements through long Chain-of-Thought (CoT). However, existing benchmarks mainly focus on immediate, single-horizon tasks, failing to adequately evaluate models' ability to understand and respond to complex, long-horizon scenarios. To address this incomplete evaluation of Large Reasoning Models (LRMs), we propose R-HORIZON, a method designed to stimulate long-horizon reasoning behaviors in LRMs through query composition. Based on R-HORIZON, we construct a long-horizon reasoning benchmark, comprising complex multi-step reasoning tasks with interdependent problems that span long reasoning horizons. Through comprehensive evaluation of LRMs using the R-HORIZON benchmark, we find that even the most advanced LRMs suffer significant performance degradation. Our analysis reveals that LRMs exhibit limited effective reasoning length and struggle to allocate thinking budget across multiple problems appropriately. Recognizing these limitations, we use R-HORIZON to construct long-horizon reasoning data for reinforcement learning with verified rewards (RLVR). Compared to training with single-horizon data, RLVR with R-HORIZON not only substantially improves performance on the multi-horizon reasoning tasks, but also promotes accuracy on standard reasoning tasks, with an increase of 7.5 on AIME2024. These results position R-HORIZON as a scalable, controllable, and low-cost paradigm for enhancing and evaluating the long-horizon reasoning capabilities of LRMs.

cs.AI↗

arXiv2025

BAPO: Stabilizing Off-Policy Reinforcement Learning for LLMs via Balanced Policy Optimization with Adaptive Clipping

Reinforcement learning (RL) has recently become the core paradigm for aligning and strengthening large language models (LLMs). Yet, applying RL in off-policy settings--where stale data from past policies are used for training--improves sample efficiency, but remains challenging: policy entropy declines sharply, optimization often becomes unstable and may even collapse. Through theoretical and empirical analysis, we identify two key insights: (i) an imbalance in optimization, where negative-advantage samples dominate the policy gradient, suppressing useful behaviors and risking gradient explosions; and (ii) the derived Entropy-Clip Rule, which reveals that the fixed clipping mechanism in PPO-like objectives systematically blocks entropy-increasing updates, thereby driving the policy toward over-exploitation at the expense of exploration. Building on these insights, we propose BAlanced Policy Optimization with Adaptive Clipping (BAPO), a simple yet effective method that dynamically adjusts clipping bounds to adaptively re-balance positive and negative contributions, preserve entropy, and stabilize RL optimization. Across diverse off-policy scenarios--including sample replay and partial rollout--BAPO achieves fast, stable, and data-efficient training. On AIME 2024 and AIME 2025 benchmarks, our 7B BAPO model surpasses open-source counterparts such as SkyWork-OR1-7B, while our 32B BAPO model not only achieves state-of-the-art results among models of the same scale but also outperforms leading proprietary systems like o3-mini and Gemini-2.5-Flash-Thinking.

cs.LG↗

arXiv2025

What's Wrong with Your Code Generated by Large Language Models? An Extensive Study

The increasing development of LLMs in code generation has drawn significant attention among researchers. To enhance LLM-based code generation ability, current efforts are predominantly directed towards collecting high-quality datasets and leveraging diverse training technologies. However, there is a notable lack of comprehensive studies examining the limitations and boundaries of existing methods. To bridge this gap, we conducted an extensive empirical study evaluating the performance of three leading closed-source LLMs and six popular open-source LLMs on three commonly used benchmarks. Our investigation, which evaluated the length, cyclomatic complexity and API number of the generated code, revealed that these LLMs face challenges in generating successful code for more complex problems, and tend to produce code that is shorter yet more complicated as compared to canonical solutions. Additionally, we developed a taxonomy of bugs for incorrect codes that includes three categories and ten sub-categories, and analyzed the root cause for common bug types. To better understand the performance of LLMs in real-world projects, we also manually created a real-world benchmark RWPB. We analyzed bugs on RWPB to highlight distinct differences in bug distributions between actual scenarios and existing benchmarks. Finally, we propose a novel training-free iterative method that introduces self-critique, enabling LLMs to critique and correct their generated code based on bug types and compiler feedback. Our comprehensive and extensive study provides insights into the current limitations of LLM-based code generation and opportunities for enhancing the accuracy and quality of the generated code.

cs.SE↗