Xuanjing Huang

Modern coding scaffolds turn LLMs into capable software agents, but their ability to follow scaffold-specified instructions remains under-examined, especially when constraints are heterogeneous and persist across interactions. To fill this gap, we introduce OctoBench, which benchmarks scaffold-aw...

Existing evaluation of Large Language Models (LLMs) on static benchmarks is vulnerable to data contamination and leaderboard overfitting, critical issues that obscure true model capabilities. To address this, we introduce LLMEval-Fair, a framework for dynamic evaluation of LLMs. LLMEval-Fair is b...

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 imbalance during this process: the model readily generat...

Vision-Language Models (VLMs) have demonstrated remarkable progress in multimodal tasks, but remain susceptible to hallucinations, where generated text deviates from the underlying visual content. Existing hallucination detection methods primarily rely on output logits or external verification to...

Reinforcement Learning (RL) in real-world environments often suffers from ambiguous or incomplete reward supervision, which undermines policy stability and generalization. Such noise may cause models to ignore key information or even collapse in advantage estimation. We find that a strong value m...

The GPT-4 technical report suggests that downstream performance can be predicted from pre-training signals, but offers little methodological detail on how to quantify this. This work address this gap by modeling knowledge retention, the capacity of a pre-trained language model to memorize factual...

Reward models (RMs) are the surrogate objectives in reinforcement learning from human feedback (RLHF), and their scores directly steer policy optimization. We show that standard RM training is vulnerable in data subsets where response quality depends only weakly on the context: such instances enc...

Generative Reward Models (GenRMs) and LLM-as-a-Judge exhibit deceptive alignment by producing correct judgments for incorrect reasons, as they are trained and evaluated to prioritize Outcome Accuracy , which undermines their ability to generalize during RLHF. We introduce Rationale Consistency ,...

Long chain-of-thought (CoT) trajectories provide rich supervision signals for distilling reasoning from teacher to student LLMs. However, both prior work and our experiments show that trajectories from stronger teachers do not necessarily yield better students, highlighting the importance of data...

Language agents, i.e., LLM agents, progress rapidly and are increasingly deployed in production environments. This trend underscores the urgent need for rigorous and realistic evaluations. However, most existing benchmarks evaluate agents in simplified, idealized settings. They typically rely on...

Instruction following refers to the ability of large language models (LLMs) to generate outputs that satisfy all specified constraints. Existing research has primarily focused on constraint categories, offering limited evaluation dimensions and little guidance for improving instruction-following...

Effective tool use is essential for large language models (LLMs) to interact with their environment. However, progress is limited by the lack of efficient reinforcement learning (RL) frameworks specifically designed for tool use, due to challenges in constructing stable training environments and...

Test-time scaling has emerged as a transformative paradigm for enhancing the performance of large reasoning models, enabling dynamic allocation of computational resources during inference. However, as the landscape of reasoning models rapidly expands, a critical question remains: how can we syste...

Previous research has shown that LLMs finetuned on incorrect completions within narrow domains ( \textit{e.g.} , insecure code or incorrect medical advice) can become broadly misaligned to exhibit harmful behaviors, which is called \textit{emergent misalignment} . In this work, we investigate whe...

Mechanistic Interpretability (MI) has emerged as a vital approach to demystify the opaque decision-making of Large Language Models (LLMs). However, existing reviews primarily treat MI as an observational science, summarizing analytical insights while lacking a systematic framework for actionable...

Tool-use capabilities are vital for Large Language Models (LLMs) in finance, a domain characterized by massive investment targets and data-intensive inquiries. However, existing data synthesis methods typically rely on a reverse synthesis paradigm, generating user queries from pre-sampled tools....

Vision–Language Models (VLMs) have demonstrated strong capabilities in tasks that require joint understanding of text and images. However, as many VLMs are built upon pre-trained large language models, they often over-rely on linguistic priors at the expense of visual features, causing persistent...

The rapid advancement of large language models (LLMs) has accelerated progress toward universal AI assistants. However, existing benchmarks for personalized assistants remain misaligned with real-world user-assistant interactions, failing to capture the complexity of external contexts and users’...

Large Language Models (LLMs) still struggle with the “lost-in-the-middle” problem, where critical information located in the middle of long-context inputs is often underrepresented or lost. While existing methods attempt to address this by combining multi-scale rotary position embeddings (RoPE),...

Recent commercial systems such as Suno demonstrate strong capabilities in long-form song generation, while academic research remains largely non-reproducible due to the lack of publicly available training data, hindering fair comparison and progress. To this end, we release a fully open-source sy...

Verifiers have been demonstrated to enhance LLM reasoning via test-time scaling (TTS). Yet, they face significant challenges in complex domains. Error propagation from incorrect intermediate reasoning can lead to false positives for seemingly plausible solutions, while lacking external grounding...

Conventional Retrieval-Augmented Generation (RAG) systems often struggle with complex multi-hop queries over long documents due to their single-pass retrieval. We introduce MM-Doc-R1 , a novel framework that employs an agentic, vision-aware workflow to address long document visual question answer...

Speech quality assessment (SQA) is typically formulated as a score regression task based on subjective ratings, such as the Mean Opinion Score (MOS), which inherently suffer from inconsistent standards and limit cross-dataset training and evaluation. To address these limitations, we reformulate S...