Skip to content
conftrace_

contrastive group relative policy optimization

1 paper

Explore in graph

Also known as

C-GRPO

Co-occurring keywords · papers mentioning both

multimodal reward model (1) embodied question answering (1) test-time scaling (1) reward model (1) embodied ai (1) multimodal learning (1) question answering (1) policy optimization (1) reinforcement learning (1) contrastive learning (1)

Papers

EQA-RM: A Generative Embodied Reward Model with Test-time Scaling EMNLP 2025