Unverified50% confidenceBenchmarkExact time
In the pre-registered experiment, the 12B model passed all 96 trials on the 24-task order-desk test suite (96/96).
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
11/11/2026
First Seen
Valid until: 11/25/2026
Sources
Related Entities
Related Claims
UnverifiedIn the pre-registered experiment, the 4B model passed 70 of the same 96 trials.78% similarUnverifiedIn a preregistered study, confidence intervals relying solely on LLM judges failed to cover the full human-annotation ground truth in 18 of 21 task-and-judge combinations.74% similarUnverifiedIn a test that used Traework to fully simulate a product development workflow, the tool retrieved more than 100 sources during the research phase, but after manual verification only about 40 were truly usable, while the other 60 needed to be discarded or re-investigated.73% similarUnverifiedThis benchmark uses 10 original short documents of 200 to 250 words each, with 18 planted anomalies, 37 scoring entries in total (18 dirty and 19 clean), testing 13 models with 74 prompts per model.73% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1161191API
curl https://kongchang.com/api/v1/knowledge/claims/1161191MCP
get_claim(id=1161191)