Unverified50% confidenceBenchmarkExact time
The developer built the ishuman benchmark, in which 2.0 was judged to be human 23.5% of the time, the base model abliterated 0.3%, the official original version 15.1%, and the base model with prompting 6.8%.
1
Sources
50%
Confidence
Long-term
Relevance
10/2/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedIn a test simulating real Cursor coding, the score improved from 34.1% to 55.5%.64% similarUnverifiedIn multiple benchmarks, even the most advanced models can have hallucination rates of 15%–30% in specialized technical domains.63% similarUnverifiedThe Intelligent Chemical Industry Large Model 3.0 Pro achieved a text Q&A accuracy rate of 81.96% and a multimodal Q&A accuracy rate of 80.75%, representing improvements of 20.2% and 31.4% respectively in overall scores compared to the previous generation.62% similarUnverifiedOn gpt-oss-120B, the conformity rate in the Asch conformity experiment under blind testing conditions was 0%, while the conformity rate rose to 83.3% when the paradigm was explicitly named.62% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/968170API
curl https://kongchang.com/api/v1/knowledge/claims/968170MCP
get_claim(id=968170)