Co-Scientist evaluation: Severe hallucinations drop to 4%, fabrication to 0%
SRSchmidgall · x · 2026-08-29
Double-blind reviews of 150 AI-generated papers showed that Co-Scientist reduced severe data hallucination from 90% to 4%, extreme data fabrication to 0%, and plagiarism from 60% to 16%, while significantly improving proper attribution.
Related event: Co-Scientist Review: Hallucinations Plummet but Limits Remain(3 posts)→
More from Models
- Image inputs now ~10,000 per dollar, but re-referencing an image doubles token cost — smtabatabaie · 2026-10-15
- "Argon doesn't trust me": researcher laments overzealous model refusal — thegautamkamath · 2026-10-15
- User finds Grok shockingly bad at finding X posts, prefers Opus with browser — JasonBotterill · 2026-10-15
- "Claude-Next" Reportedly Returns to Claude CLI and Web via Stealth Opus Routing — koltregaskes · 2026-10-15
- 'Thinking leakage': no-think RL gains in hybrid models just reuse think-mode skills — kaust-generative-ai · 2026-10-15
- Cline: Gemini 4 Argon saw 3x Claude's day-one adoption and 30% fewer tokens — JMateosGarcia · 2026-10-15