6 papers
Composition Collapse: Stable Factual Knowledge Does Not Imply Compositional Reasoning
Zhe Yu, Wenpeng Xing, Yunzhao Wei +4
Post-training is routinely evaluated through aggregate benchmark scores that treat multi-hop reasoning as a single capability -- as if a model that answers more questions correctly…
ScientistOne: Towards Human-Level Autonomous Research via Chain-of-Evidence
Rui Meng, Bhavana Dalvi Mishra, Jiefeng Chen +10
Autonomous research agents produce competitive solutions and professional-looking manuscripts, yet their outputs contain verifiability failures undetectable by surface-level evalua…
SETS: Leveraging Self-Verification and Self-Correction for Improved Test-Time Scaling
Jiefeng Chen, Jie Ren, Xinyun Chen +4
Recent advancements in Large Language Models (LLMs) have created new opportunities to enhance performance on complex reasoning tasks by leveraging test-time computation. However, e…
MLE-STAR: Machine Learning Engineering Agent via Search and Targeted Refinement
Jaehyun Nam, Jinsung Yoon, Jiefeng Chen +3
Agents based on large language models (LLMs) for machine learning engineering (MLE) can automatically implement ML models via code generation. However, existing approaches to build…
DynScaling: Efficient Verifier-free Inference Scaling via Dynamic and Integrated Sampling
Fei Wang, Xingchen Wan, Ruoxi Sun +2
Inference-time scaling has proven effective in boosting large language model (LLM) performance through increased test-time computation. Yet, its practical application is often hind…
Astute RAG: Overcoming Imperfect Retrieval Augmentation and Knowledge Conflicts for Large Language Models
Fei Wang, Xingchen Wan, Ruoxi Sun +2
Retrieval augmented generation (RAG), while effectively integrating external knowledge to address the inherent limitations of large language models (LLMs), can be hindered by imper…