Unverified50% confidenceFactExact time
Noul questions in Jev have the model return a floating-point number between 0 and 1, representing the confidence that a statement is true.
1
Sources
50%
Confidence
Long-term
Relevance
9/22/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn the first round of 34 logical reasoning questions, a single-head model similar to JEV achieved an accuracy of 94.12%, while GAN8 achieved 93.01%, nearly tied.67% similarUnverifiedThe experiment compared Jev in zero-shot and few-shot settings against 11 classical machine learning models.64% 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.62% 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.61% similarUnverifiedA fusion strategy that prioritizes JEV and only hands off when confidence is below 0.7 collapses into JEV, with approximately 70% of questions never leaving JEV and an ECE of approximately 0.25.60% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/939730API
curl https://kongchang.com/api/v1/knowledge/claims/939730MCP
get_claim(id=939730)