Unverified50% confidenceOpinionExact time
Jia Xiaosong believes imitation learning can take a model from 0 points to 90 points and reinforcement learning can tackle the remaining 10 points, and the two are not substitutes for each other; reinforcement learning starting completely from scratch converges extremely slowly.
1
Sources
50%
Confidence
Long-term
Relevance
10/19/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedKnowledge distillation has the student network learn the soft probability distribution output by the teacher network, enabling the student model to retain core reasoning capabilities while reducing its parameter count by more than 10 times.67% similarUnverifiedThe 9-billion-parameter model in the paper lagged 6 points behind a clean summarization baseline before being trained with reinforcement learning.66% similarUnverifiedThrough distillation, student models with only one-tenth the parameters of the teacher model can often retain 80%–95% of the teacher model's performance.65% similarUnverified当不动动作占据80%以上帧时，模型通过最小化交叉熵损失可在不学习有意义特征的情况下达到80%准确率，这称为多数类捷径65% similarUnverifiedThe experiment compared Jev in zero-shot and few-shot settings against 11 classical machine learning models.64% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1065553API
curl https://kongchang.com/api/v1/knowledge/claims/1065553MCP
get_claim(id=1065553)