Unverified50% confidenceBenchmarkExact time
The best configuration of the question answering approach achieved an LLM annotation accuracy of 60.5%.
1
Sources
50%
Confidence
Long-term
Relevance
9/24/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedGPT-5.4 has an accuracy of 50.5%, while its paired always-predict-up baseline is 51.6%.63% similarUnverified一张包含30个字段、单字段准确率98%的发票，完全正确的概率约为54.5%62% similarUnverifiedThe framework achieves an overall accuracy improvement of 6.9% over the current state of the art on the MERLIN benchmark.60% similarUnverifiedPrecision Mode leads the strongest frontier model chunk-and-merge baseline solution by 7 percentage points in accuracy.59% similarUnverifiedTop-1% accuracy is a metric for evaluating the output consistency between a quantized model and the original full-precision model, focusing on whether the top 1% of tokens with the highest model confidence are consistent with the full-precision version.59% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/946148API
curl https://kongchang.com/api/v1/knowledge/claims/946148MCP
get_claim(id=946148)