Unverified50% confidenceFactExact time
The LAD (Locality-Aware Distillation) module restricts knowledge transfer to the student's failure trajectories, and only to problems on which the teacher performs better.
1
Sources
50%
Confidence
Long-term
Relevance
11/10/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWeak-to-strong generalization does not hold universally, but depends on conditions such as target structure and training methodology; when conditions are not met, the student may fail to surpass the teacher69% similarUnverifiedThe teacher and anchor dimensions must work in concert: having only a high-quality teacher without an anchor model may lead to gradual drift, while having only an anchor but poor-quality teacher pseudo-labels makes it difficult to leverage the massive amount of client-side data.68% similarUnverifiedMihail Eric observed that if students can get a coding agent to complete coursework with a single prompt, traditional teaching methods have already failed.68% similarUnverifiedTools replace the execution step, not the understanding step; once understanding is outsourced, learners will lose the ability to judge whether a tool's output is right or wrong.67% similarUnverifiedIn the skill routing scenario, routing interference and catastrophic forgetting in class-incremental learning are homologous in problem structure.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1154303API
curl https://kongchang.com/api/v1/knowledge/claims/1154303MCP
get_claim(id=1154303)