Feng Chen

7 papers Microsoft AI / Stanford University

• ACT as Human: Multimodal Large Language Model Data Annotation with Critical Thinking
• Adaptive Fission: Post-training Encoding for Low-latency Spike Neural Networks
• Alternating Gradient Flows: A Theory of Feature Learning in Two-layer Neural Networks
• EvaLearn: Quantifying the Learning Capability and Efficiency of LLMs via Sequential Problem Solving
• Informed Correctors for Discrete Diffusion Models
• Rethinking Fine-Tuning when Scaling Test-Time Compute: Limiting Confidence Improves Mathematical Reasoning
• SolverLLM: Leveraging Test-Time Scaling for Optimization Problem via LLM-Guided Search