Unverified50% confidenceFactExact time
After removing context, the accuracy gap of the base models on the legal benchmark can reach up to 27 percentage points.
1
Sources
50%
Confidence
Long-term
Relevance
11/6/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedUnder the influence of multi-turn narratives, the models' final judgments shifted by an average of 25 percentage points compared to matched single-turn baselines.68% similarUnverifiedIn the most difficult single-concept quantitative category, most models' accuracy falls below 30%68% similarUnverifiedWhen single-step accuracy is 96%, the system success rate after 30 chained steps is about 29%.67% similarUnverifiedEven the best-performing models can only recover 30% of the evaluated evidence and 25% of the limitations descriptions.66% similarUnverifiedWhen models were provided with complete company records, strict accuracy actually dropped from 41% to 18%, while recall rose and precision collapsed.66% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1141675API
curl https://kongchang.com/api/v1/knowledge/claims/1141675MCP
get_claim(id=1141675)