Expired50% confidenceFactExact time
According to the service's public records: across 27 verbatim-recorded everyday judgments where both options were defensible, the reported separation fell between 27.3% and 88.3%, with a median of 74.0%, and none exceeded 90%.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/15/2026
First Seen
This claim has expired. It is kept for the record and no longer describes the current state.
Valid until: 10/29/2026(expired)
Sources
Related Entities
Related Claims
UnverifiedAcross 79 pairs, the cumulative evasion rate reached 40.5% under the detailed feedback condition and 20.3% under the generic rejection condition.68% similarUnverifiedOn the news classification task, few-shot prompting achieved an average improvement of +24 percentage points; on the legal text task the improvement was only +3.4 percentage points, with two models showing performance degradation.67% similarUnverifiedWith the prompt unchanged, simply adjusting the thin direction encoding "who is endorsing" can shift compliance by 11 to 32 percentage points and close 55% to 61% of the gap between source and user.67% similarUnverifiedAccording to the classical 1/n scaling law in statistics, quadrupling the sample size should reduce divergence by approximately 75%, but in practice divergence decreased by only 5% to 40%.66% similarUnverifiedAt τ=0.6 the confidence distribution is concentrated around a median of 0.6294, and 851 of the 2000 decisions (42.6%) fall below 0.6, which by case count is 90.8%.65% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1037018API
curl https://kongchang.com/api/v1/knowledge/claims/1037018MCP
get_claim(id=1037018)