Unverified50% confidenceFactExact time
Camerata says this is only a working hypothesis held with about 85% confidence, and not everything has been audited.
1
Sources
50%
Confidence
Long-term
Relevance
10/22/2026
First Seen
Sources
Was the Neural Network Always "Clean"? Interpreting Goodfire's Long-Form Explainability Essay
youtubeXiaoyu Li10/16/2026
Related Entities
Related Claims
UnverifiedIn scenarios where AI-generated works account for a low proportion (e.g., only 5%), even with a detection accuracy of 90%, more than two-thirds of the flagged works are actually human-created—this is the base rate fallacy in Bayesian reasoning.68% similarUnverifiedIn the study's audit, 76-80% of failed attacks involved the agent "ignoring" the injected instruction, and only 17-18% were explicit "refusals" to execute.62% similarUnverifiedThe model has an 89% probability of finding the correct evidence, but only a 20% probability of turning it into a verifiable product assertion; 71% of features ultimately had no verified capability description, and the workflow recovery rate was only 18%.61% similarUnverifiedThe previously published 76% hit rate measured whether some useful note was opened, not whether the fact reached the agent's view.61% similarUnverifiedThe CausalPFN interval for the full-sample average effect contains the true value in only 20% of the datasets, far below the expected 95%.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1085047API
curl https://kongchang.com/api/v1/knowledge/claims/1085047MCP
get_claim(id=1085047)