Loading activity…
@_xjdr
building AI that wont embarrass me in front of my own standards
AI
A user names DSV4.1 Flash and GLM 5.3 among the models that currently interest them.
AI
David Pfau objected to the definition, while Dylan Hadfield-Menell called the evaluation metric optimization.
AI
OpenAI says it drew on an independent mathematics advisory group’s advice about how to release the results.
AI
Underdog says its Mac assistant stores personal data on-device, with iOS, Windows and Android versions planned.
AI
A user jokes about making an “xgboost+MAB in a trenchcoat” model and calling it something like “stevia.”
AI
A user credits fast-jev-compaction, which they describe as a Claude plugin that uses the Jev model to review unnecessary tool calls.
AI
One user wonders whether the model approach is making a return, greeting the idea with a brief “cool.”
AI
Pseudonymous developer xjdr shares AI summary from research journal and git logs on Entropix.
AI
Claude's official account confirms the launch rates stay in effect indefinitely.
AI
Announces shift to advisor role and return to posting on the platform.
AI
AI experts react on social media to Axios report on withheld evaluation criteria.
AI
Internal Astra model produces Lean-certified proofs for ten open problems across math and theoretical computer science.
AI
OpenAI engineers show retained reasoning and compaction lift base model results on complex tasks.
AI
Community replies urge larger base models, tool calling, and greater openness.
