Kehua Feng

I-Chun Chern, Steffi Chern, Shiqi Chen et al. · cmu

The emergence of generative pre-trained models has facilitated the synthesis of high-quality text, but it has also posed challenges in identifying factual errors in the generated text. In particular: (1) A wider range of tasks now face an increasing risk of containing factual errors when handled by generative models. (2) Generated texts tend to be lengthy and lack a clearly defined granularity for individual facts. (3) There is a scarcity of explicit evidence available during the process of fact checking. With the above challenges in mind, in this paper, we propose FacTool, a task and domain agnostic framework for detecting factual errors of texts generated by large language models (e.g., ChatGPT). Experiments on four different tasks (knowledge-based QA, code generation, mathematical reasoning, and scientific literature review) show the efficacy of the proposed method. We release the code of FacTool associated with ChatGPT plugin interface at https://github.com/GAIR-NLP/factool .

Kehua Feng, Keyan Ding, Hongzhi Tan et al.

Reliable evaluation of large language models (LLMs) is impeded by two key challenges: objective metrics often fail to reflect human perception of natural language, and exhaustive human labeling is prohibitively expensive. Here, we propose a sample-efficient human evaluation method for LLMs based on the principle of MAximum Discrepancy (MAD) Competition. Our method automatically and adaptively selects a compact set of input instructions that maximize semantic discrepancy between pairs of LLM responses. Human evaluators then perform three-alternative forced choices on these paired responses, which are aggregated into a global ranking using Elo rating. We apply our approach to compare eight widely used LLMs across four tasks: scientific knowledge understanding, mathematical reasoning, creative and functional writing, and code generation and explanation. Experimental results show that our sample-efficient evaluation method recovers "gold-standard" model rankings with a handful of MAD-selected instructions, reveals respective strengths and weaknesses of each LLM, and offers nuanced insights to guide future LLM development. Code is available at https://github.com/weiji-Feng/MAD-Eval .

Xiang Zhuang, Chenyi Zhou, Kehua Feng et al.

Artificial intelligence has demonstrated remarkable capability in predicting scientific properties, yet scientific discovery remains an inherently physical, long-horizon pursuit governed by experimental cycles. Most current computational approaches are misaligned with this reality, framing discovery as isolated, task-specific predictions rather than continuous interaction with the physical world. Here, we argue for embodied science, a paradigm that reframes scientific discovery as a closed loop tightly coupling agentic reasoning with physical execution. We propose a unified Perception-Language-Action-Discovery (PLAD) framework, wherein embodied agents perceive experimental environments, reason over scientific knowledge, execute physical interventions, and internalize outcomes to drive subsequent exploration. By grounding computational reasoning in robust physical feedback, this approach bridges the gap between digital prediction and empirical validation, offering a roadmap for autonomous discovery systems in the life and chemical sciences.

Mingyang Rao, Kehua Feng, Zhihui Zhu et al.

While Large Language Models (LLMs) have revolutionized scientific text processing, they exhibit a significant capability gap when interpreting chemical reaction diagrams. We identify two fundamental bottlenecks restricting current systems: a Visual Deficit, where generic vision encoders struggle to resolve the strict topological connectivity of dense molecular graphs, and a Semantic Disconnect, where standard linear strings, such as SMILES, fail to effectively activate the model's latent chemical reasoning. To bridge these gaps, we propose the Chemical Visual Activation (ChemVA) framework, which employs a Visual Anchor mechanism to ground functional groups via hybrid-granularity detection, followed by a semantic alignment approach that translates visual features into entity names to maximize knowledge activation in LLMs. We evaluate our approach on OCRD-Bench, a newly constructed dataset featuring dense visual-semantic contexts and comprehensive reaction coverage to evaluate the full spectrum from recognition to reasoning. Extensive experiments on OCRD-Bench demonstrate that ChemVA achieves 92.0% structural recognition accuracy. By bridging visual and semantic bottlenecks, our framework delivers a consistent performance gain of approximately 20 percentage points across 9 diverse LLMs, enabling open-weight models to rival proprietary SOTA systems in complex chemical reasoning tasks.

Shunyang Luo, Peibei Cao, Zhihui Zhu et al.

Reward models (RMs) are central to aligning large language models, yet their practical effectiveness hinges on generalization to unseen prompts and shifting distributions. Most existing RM evaluations rely on static, pre-annotated preference datasets, which provide limited coverage and often fail to faithfully assess generalization in open-world settings. We introduce Pairwise Maximum Discrepancy Competition (PMDC), a dynamic and annotation-efficient framework for evaluating RM generalization using a large, unlabeled, open-domain prompt pool. PMDC actively selects prompt--response pairs that maximize disagreement between two RMs, yielding a compact set of highly contentious test cases. These cases are adjudicated by an oracle, and the resulting outcomes are aggregated via a Bradley--Terry model to produce a global ranking and pairwise win-rate landscape of RMs. We apply PMDC to re-evaluate 10 representative RMs and observe substantial rank reshuffling compared with conventional benchmarks. Qualitative analyses further uncover systematic generalization failures, providing valuable insights for improving reward modeling.

Yuqi Tang, Jing Yu, Zichang Su et al.

Clinical diagnosis begins with doctor-patient interaction, during which physicians iteratively gather information, determine examination and refine differential diagnosis through patients' response. This dynamic clinical-reasoning process is poorly represented by existing LLM benchmarks that focus on static question-answering. To mitigate these gaps, recent methods explore dynamic medical frameworks involving interactive clinical dialogues. Although effective, they often rely on limited, contamination-prone datasets and lack granular, multi-level evaluation. In this work, we propose ClinDEF, a dynamic framework for assessing clinical reasoning in LLMs through simulated diagnostic dialogues. Grounded in a disease knowledge graph, our method dynamically generates patient cases and facilitates multi-turn interactions between an LLM-based doctor and an automated patient agent. Our evaluation protocol goes beyond diagnostic accuracy by incorporating fine-grained efficiency analysis and rubric-based assessment of diagnostic quality. Experiments show that ClinDEF effectively exposes critical clinical reasoning gaps in state-of-the-art LLMs, offering a more nuanced and clinically meaningful evaluation paradigm.

Kehua Feng, Keyan Ding, Yuhao Wang et al.

Recent advancements in large language models (LLMs) have accelerated progress toward artificial general intelligence, yet their potential to generate harmful content poses critical safety challenges. Existing alignment methods often struggle to cover diverse safety scenarios and remain vulnerable to adversarial attacks. In this work, we propose SAFER, a framework for Safety Alignment via eFficient Ex-Ante Reasoning. Our approach instantiates structured Ex-Ante reasoning through initial assessment, rule verification, and path calibration, and embeds predefined safety rules to provide transparent and verifiable safety judgments. Specifically, our approach consists of two training stages: (1) supervised fine-tuning with synthetic traces to teach the multi-stage Ex-Ante reasoning, and (2) step-level reasoning preference optimization to jointly enhance safety, utility, and efficiency. Experiments on multiple open-source LLMs demonstrate that SAFER significantly enhances safety performance while maintaining helpfulness and response efficiency.

Xiang Zhuang, Bin Wu, Jiyu Cui et al.

Molecular structure elucidation involves deducing a molecule's structure from various types of spectral data, which is crucial in chemical experimental analysis. While large language models (LLMs) have shown remarkable proficiency in analyzing and reasoning through complex tasks, they still encounter substantial challenges in molecular structure elucidation. We identify that these challenges largely stem from LLMs' limited grasp of specialized chemical knowledge. In this work, we introduce a Knowledge-enhanced reasoning framework for Molecular Structure Elucidation (K-MSE), leveraging Monte Carlo Tree Search for test-time scaling as a plugin. Specifically, we construct an external molecular substructure knowledge base to extend the LLMs' coverage of the chemical structure space. Furthermore, we design a specialized molecule-spectrum scorer to act as a reward model for the reasoning process, addressing the issue of inaccurate solution evaluation in LLMs. Experimental results show that our approach significantly boosts performance, particularly gaining more than 20% improvement on both GPT-4o-mini and GPT-4o. Our code is available at https://github.com/HICAI-ZJU/K-MSE.

Kehua Feng, Xinyi Shen, Weijie Wang et al.

Large language models (LLMs) are playing an increasingly important role in scientific research, yet there remains a lack of comprehensive benchmarks to evaluate the breadth and depth of scientific knowledge embedded in these models. To address this gap, we introduce SciKnowEval, a large-scale dataset designed to systematically assess LLMs across five progressive levels of scientific understanding: memory, comprehension, reasoning, discernment, and application. SciKnowEval comprises 28K multi-level questions and solutions spanning biology, chemistry, physics, and materials science. Using this benchmark, we evaluate 20 leading open-source and proprietary LLMs. The results show that while proprietary models often achieve state-of-the-art performance, substantial challenges remain -- particularly in scientific reasoning and real-world application. We envision SciKnowEval as a standard benchmark for evaluating scientific capabilities in LLMs and as a catalyst for advancing more capable and reliable scientific language models.

Yuhao Wang, Keyan Ding, Kehua Feng et al.

Protein language models have emerged as powerful tools for sequence generation, offering substantial advantages in functional optimization and denovo design. However, these models also present significant risks of generating harmful protein sequences, such as those that enhance viral transmissibility or evade immune responses. These concerns underscore critical biosafety and ethical challenges. To address these issues, we propose a Knowledge-guided Preference Optimization (KPO) framework that integrates prior knowledge via a Protein Safety Knowledge Graph. This framework utilizes an efficient graph pruning strategy to identify preferred sequences and employs reinforcement learning to minimize the risk of generating harmful proteins. Experimental results demonstrate that KPO effectively reduces the likelihood of producing hazardous sequences while maintaining high functionality, offering a robust safety assurance framework for applying generative models in biotechnology.

Kehua Feng, Keyan Ding, Zhihui Zhu et al.

While chain-of-thought (CoT) distillation from advanced large language models (LLMs) has proven effective in general reasoning tasks, it struggles in scientific domains where even advanced models often produce incorrect or superficial reasoning due to high complexity and specialized knowledge requirements. Directly distilling from such flawed outputs results in low-quality training data and limits the performance of smaller student models. To overcome this, we propose CoT-Evo, an evolutionary CoT distillation framework. It begins by constructing a diverse pool of reasoning trajectories from multiple LLM thinkers, enriches them with automatically retrieved domain knowledge, and iteratively refines the trajectories using novelty-driven selection, reflective recombination and mutation. The refinement is guided by a fitness function that evaluates answer correctness, coherence, and effective knowledge utilization. This results in a high-quality CoT dataset tailored for scientific reasoning. We employ this evolved dataset to fine-tune a compact model, which achieves state-of-the-art performance on scientific reasoning benchmarks. Our work establishes a scalable approach to synthesizing high-fidelity scientific reasoning data from diverse and fallible LLMs.

Yuqi Tang, Kehua Feng, Yunfeng Wang et al.

Evaluating the conversational abilities of large language models (LLMs) remains a challenging task. Current mainstream approaches primarily rely on the "LLM-as-a-judge" paradigm, where an LLM is prompted to serve as an evaluator to assess dialogue quality. However, such methods often suffer from various biases, which undermine the reliability and consistency of the evaluation results. To mitigate these biases, recent methods employ multiple LLMs as judges and aggregate their judgments to select the optimal assessment. Although effective, this multi-judge approach incurs significant computational overhead during inference. In this paper, we propose an efficient multi-turn dialogue evaluator that captures the collective wisdom of multiple LLM judges by aggregating their preference knowledge into a single model. Our approach preserves the advantages of diverse multi-judge feedback while drastically reducing the evaluation cost, enabling fast and flexible dialogue quality assessment. Extensive experiments on seven single rating and pairwise comparison dialogue evaluation benchmarks demonstrate that our method outperforms existing baselines across diverse scenarios, showcasing its efficiency and robustness.

Jing Yu, Yuqi Tang, Kehua Feng et al.

Large Language Models (LLMs) have shown impressive capabilities in contextual understanding and reasoning. However, evaluating their performance across diverse scientific domains remains underexplored, as existing benchmarks primarily focus on general domains and fail to capture the intricate complexity of scientific data. To bridge this gap, we construct SciCUEval, a comprehensive benchmark dataset tailored to assess the scientific context understanding capability of LLMs. It comprises ten domain-specific sub-datasets spanning biology, chemistry, physics, biomedicine, and materials science, integrating diverse data modalities including structured tables, knowledge graphs, and unstructured texts. SciCUEval systematically evaluates four core competencies: Relevant information identification, Information-absence detection, Multi-source information integration, and Context-aware inference, through a variety of question formats. We conduct extensive evaluations of state-of-the-art LLMs on SciCUEval, providing a fine-grained analysis of their strengths and limitations in scientific context understanding, and offering valuable insights for the future development of scientific-domain LLMs.

Qiang Zhang, Keyang Ding, Tianwen Lyv et al.

Large Language Models (LLMs) have emerged as a transformative power in enhancing natural language comprehension, representing a significant stride toward artificial general intelligence. The application of LLMs extends beyond conventional linguistic boundaries, encompassing specialized linguistic systems developed within various scientific disciplines. This growing interest has led to the advent of scientific LLMs, a novel subclass specifically engineered for facilitating scientific discovery. As a burgeoning area in the community of AI for Science, scientific LLMs warrant comprehensive exploration. However, a systematic and up-to-date survey introducing them is currently lacking. In this paper, we endeavor to methodically delineate the concept of "scientific language", whilst providing a thorough review of the latest advancements in scientific LLMs. Given the expansive realm of scientific disciplines, our analysis adopts a focused lens, concentrating on the biological and chemical domains. This includes an in-depth examination of LLMs for textual knowledge, small molecules, macromolecular proteins, genomic sequences, and their combinations, analyzing them in terms of model architectures, capabilities, datasets, and evaluation. Finally, we critically examine the prevailing challenges and point out promising research directions along with the advances of LLMs. By offering a comprehensive overview of technical developments in this field, this survey aspires to be an invaluable resource for researchers navigating the intricate landscape of scientific LLMs.