Unverified50% confidenceBenchmarkExact time
Raw drift magnitude cannot predict few-shot success or failure (correlation coefficient r=0.20), whereas content delta can effectively predict it (Spearman correlation rho=+0.65, p=0.043).
1
Sources
50%
Confidence
Long-term
Relevance
9/16/2026
First Seen
Sources
Related Entities
Related Claims
Unverified当模型对某个训练样本的预测概率接近1时，该样本对应的交叉熵损失趋近于零，模型几乎无法获得梯度更新信号68% similarUnverified当预测的前景像素数量很少或几乎为零时，Dice损失分母接近于零，导致损失值和梯度都可能出现剧烈震荡64% similarUnverifiedOpenAI's chart shows that at lower test-time compute levels, the model's success rate on this problem is close to zero, with occasional success only under extremely long reasoning chains.63% similarUnverifiedIf a difficult problem fails across all samples (pass@k ≈ 0), the reward variance is zero, the gradient signal nearly vanishes, and the parameter update approaches zero.63% similarUnverifiedUnder the assumption that the true mapping is uniformly random, the paper proves a lower bound on the error rate of E ≥ (M/N)δ*(B/M) + (1-M/N)(1-1/K).63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/924780API
curl https://kongchang.com/api/v1/knowledge/claims/924780MCP
get_claim(id=924780)