Unverified50% confidenceDecision RuleExact time
The v4 scoring protocol scores each of the 108 items independently; missing, duplicate, invalid, or incorrect answers are penalized only for that item's own score, but if the entire batch of JSON fails to parse, all six items in that batch cannot be scored.
1
Sources
50%
Confidence
Long-term
Relevance
10/12/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedAn LLM review panel that only examined submitted code and reported scores missed 33 out of 505 confirmed cheating cases (6.5%).63% similarUnverifiedOf the 12 responses containing content not present in the document, 5 contained planted errors, and all 5 retained the errors without correcting them.58% similarUnverifiedTomasuLLM produces zero false accepts across 4,010 audited commit-validation records.58% similarUnverifiedContradictory rules caused four legitimate refunds that should have been processed to not be issued.56% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1003832API
curl https://kongchang.com/api/v1/knowledge/claims/1003832MCP
get_claim(id=1003832)