PhyCritic：面向物理 AI 的多模态评判模型
PhyCritic: Multimodal Critic Models for Physical AI
自参照评判（critic 先自行生成预测作为内部参照再打分）这一机制可直接迁移到视觉生成/编辑结果的自动评测与偏好数据筛选，为扩散模型或 VLM 的 RLHF/RLVR 提供更稳定的奖励信号。
原文依据 the critic generates its own prediction as an internal reference before judging candidate responses
摘要译文
由模型翻译 · 术语统一 · 请对照原文核对关键结论随着大型多模态模型的快速发展，可靠的评判与批评模型已成为开放式评估和偏好对齐的关键，它们能够提供成对偏好、数值评分以及解释性理由，用以评估模型生成的回答。然而，现有批评模型主要在图像描述或图像问答等通用视觉领域训练，对涉及感知、因果推理与规划的物理 AI 任务仍鲜有探索。我们提出 PhyCritic，一个通过两阶段 RLVR 流程针对物理 AI 优化的多模态批评模型：第一阶段为物理技能预热，增强面向物理的感知与推理能力；第二阶段为自参照批评微调，批评模型在评判候选回答之前先生成自己的预测作为内部参照，从而提升判断稳定性与物理正确性。在物理类与通用多模态评判基准上，PhyCritic 相较开源基线取得显著性能提升；当其作为策略模型使用时，还能进一步改善物理落地任务中的感知与推理能力。
英文原文摘要 1168 字符
With the rapid development of large multimodal models, reliable judge and critic models have become essential for open-ended evaluation and preference alignment, providing pairwise preferences, numerical scores, and explanatory justifications for assessing model-generated responses. However, existing critics are primarily trained in general visual domains such as captioning or image question answering, leaving physical AI tasks involving perception, causal reasoning, and planning largely underexplored. We introduce PhyCritic, a multimodal critic model optimized for physical AI through a two-stage RLVR pipeline: a physical skill warmup stage that enhances physically oriented perception and reasoning, followed by self-referential critic finetuning, where the critic generates its own prediction as an internal reference before judging candidate responses, improving judgment stability and physical correctness. Across both physical and general-purpose multimodal judge benchmarks, PhyCritic achieves strong performance gains over open-source baselines and, when applied as a policy model, further improves perception and reasoning in physically grounded tasks.