这篇论文给多模态蒸馏提了个新招：用师生分歧定位该纠的感知错误，八个基准分数明显涨，做蒸馏的可以看。
论文提出感知纠正蒸馏（PCD），用于多模态推理器的感知蒸馏。PCD 通过下游失败和师生分歧两个见证信号的乘积来识别可纠正的感知错误，并保持推理目标不变。在八个基准上，PCD 将 8B→2B 蒸馏的宏平均从 44.50 提升到 47.28，将 32B→8B 从 56.94 提升到 61.22。2B 消融实验中，去掉 PCD 使留出集平均下降 2.22 分，去掉分离式 rollout 下降 0.88 分。
Correcting What You Cannot See: Credit Assignment for Perception Distillation in Multimodal Reasoners
On-policy distillation provides dense supervision for multimodal reasoners, but its trajectory-level reward cannot determine whether a failed answer arose from perception or subsequent reasoning. Perception Success Rate (PSR), estimated from multiple reasonings sharing one perception, remains ambiguous because low success conflates perceptual insufficiency with reasoning difficulty. We introduce \textbf{Perception-Correction Distillation (PCD)}, a label-free method that identifies correctable perception failures using downstream failure and teacher--student disagreement as complementary witnesses. Their product, , forms a soft AND gate that strengthens distillation only when both witnesses are present. We motivate this rule through Bayesian evidence combination and show that multiplication is the unique normalized bilinear gate that vanishes when either witness is absent. PCD uses separated perception--reasoning rollouts and mean-preserving weights, leaving the reasoning objective unchanged. Across eight benchmarks, PCD improves the 8B 2B macro average from 44.50 with OPD to 47.28 and the 32B 8B result from 56.94 to 61.22. In matched 2B ablations, removing PCD and separated rollout reduces held-out average by 2.22 and 0.88 points, respectively. Effective multimodal distillation therefore depends not only on what the teacher predicts, but also on identifying when perception is the appropriate target of correction.