Unverified50% confidenceEventExact time
In a test on October 14, 2026, an HR request text containing information about a colleague's sick leave after back surgery was rejected by the local model all 8 times it was sent (HTTP 400), while all 8 sends not containing that information passed (HTTP 204), and no two of the 8 rejection reasons given were exactly the same.
1
Sources
50%
Confidence
Short-term (~14 days)
Relevance
10/14/2026
First Seen
Valid until: 10/28/2026
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%).58% similarUnverifiedOf the 12 responses containing content not present in the document, 5 contained planted errors, and all 5 retained the errors without correcting them.54% similarUnverified8 of the 9 adapters failed to pass the gate, with failure reasons including data contamination, over-refusal, and even one adapter denying a real court case.54% similarUnverifiedThe 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.53% similarUnverifiedOf the 43 decisions in the ledger, 35 were rejected, most of them malformed credentials.52% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1025544API
curl https://kongchang.com/api/v1/knowledge/claims/1025544MCP
get_claim(id=1025544)