Unverified50% confidenceFactExact time
Steered-Teacher OPSD (ST-OPSD) simultaneously mitigates both failure modes of reference hallucination and reconsideration suppression through activation steering techniques.
1
Sources
50%
Confidence
Long-term
Relevance
10/27/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedExperimental results show that SAPD improves performance under different teacher configurations and effectively mitigates catastrophic forgetting compared with SFT.69% similarUnverifiedOPRD's design idea is to give reminder signals during training and completely remove the reminders at inference time, so that the student model ultimately learns to complete long-horizon tasks independently without reminders.68% similarUnverifiedThe key mechanism of RM-OPD is applying token-level dense supervision on critiques generated by the student itself, making the supervision distribution consistent with the student's current policy distribution and alleviating the training-inference distribution mismatch problem.67% similarVerifiedOnline policy distillation (OPD) has the student model generate its own training responses, which are then supervised by the teacher model.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1102389API
curl https://kongchang.com/api/v1/knowledge/claims/1102389MCP
get_claim(id=1102389)