Benyou Wang

We propose a new distillation approach that removes the input question for adaptive and efficient reasoning.

Unsupervised Prefix Fine-Tuning Method for Reasoning Models

This paper introduce CoRT, a post-training framework for teaching large reasoning LLMs to leverage CI effectively and efficiently.

We introduce TwinMarket, an LLM-based multi-agent framework that simulates socio-economic systems by modeling how individual behaviors interact to produce emergent market dynamics like bubbles and crashes.