PC-Dr.DPO improves reliability of majority-vote pseudo-labels in test-time RL
Prompt-Conditional Distributionally Robust DPO makes updates from sampling-sensitive prompts more conservative.
1 stories, newest first · one card per story
Prompt-Conditional Distributionally Robust DPO makes updates from sampling-sensitive prompts more conservative.
Today 1.6K readers · 2.5K views · 780 reads · all time 189K views