Siyuan Qi

Mindmaster Roleplay: A Social Reasoning and Planning Benchmark

Beyond Next-Token Prediction: Diffusion vs. Autoregressive Reasoning in LLMs

In-Context Editing: Learning Knowledge from Self-Induced Distributions

MMKE-Bench: A Multimodal Editing Benchmark for Diverse Visual Knowledge

Federated Coordination: Private and Distributed Strategy Alignment

The Critic as an Explorer: Lightweight and Provably Efficient Exploration for Deep Reinforcement Learning