[Bug] Agentic reasoning produces inferior outputs compared to alternative models on fact-checking tasks
Open 💬 0 comments Opened Jul 10, 2026 by neoncapy
Bug Description
Fable here is doing a worse job over and over than Sol from chatgbt with chatgbt's Sol having to correct it at fact checking an email
Environment Info
- Platform: darwin
- Terminal: vscode
- Version: 2.1.206
- Feedback ID: c956ac67-64b8-456e-92fd-27a27b8a0d18
Errors
[]