Unverified50% confidenceFactExact time
The research team collected 103,939 scored model responses across ten configurations: eight LLMs with reasoning disabled plus two of those with maximum reasoning capability enabled.
1
Sources
50%
Confidence
Long-term
Relevance
10/8/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe paper designed an experiment with 9,000 calls to compare representational consensus against paraphrase consensus.71% similarUnverifiedThe research team conducted controlled experiments on four trained DeBERTa checkpoints across a total of 7,890 claims.70% similarUnverifiedThe evaluation covered nine text-based scientific benchmarks, sampling 4,531 questions and 100 matched profiles, with 4,488 items completing all five conditions in full.69% similarUnverifiedThe researchers constructed a benchmark test set called ALIGNMENTFORECASTBENCH, covering more than 5,000 prediction questions, 17 target models, 32 datasets, and 16 failure modes.69% similarUnverifiedThe evaluation comprised 108 checks across three major modules: backend API and full pipeline at 65 points, frontend pages and real interaction at 25 points, and build and packaging at 10 points.68% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/986643API
curl https://kongchang.com/api/v1/knowledge/claims/986643MCP
get_claim(id=986643)