Picture for Myeongjun Oh

Myeongjun Oh

Parallel Tempering Initial Sampling in Inference-Time Reward Alignment

Add code
May 29, 2026
Viaarxiv icon