A rollout-conditioned contrastive distillation loss plus sparse anchored cross-entropy improves injected-knowledge accuracy in MLLMs while keeping retention close to the base model.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
RoCo-ACE: Rollout-Conditioned Online Distillation for Retention-Aware Knowledge Injection
A rollout-conditioned contrastive distillation loss plus sparse anchored cross-entropy improves injected-knowledge accuracy in MLLMs while keeping retention close to the base model.