return2ozma@lemmy.world to Technology@lemmy.worldEnglish · 21 hours agoAnthropic says its latest AI model is too powerful for public release and that it broke containment during testingwww.businessinsider.comexternal-linkmessage-square103fedilinkarrow-up1210
arrow-up1210external-linkAnthropic says its latest AI model is too powerful for public release and that it broke containment during testingwww.businessinsider.comreturn2ozma@lemmy.world to Technology@lemmy.worldEnglish · 21 hours agomessage-square103fedilink
minus-squaretleb@lemmy.calinkfedilinkEnglisharrow-up4·19 hours agoThere was some references to experimental models not publicly available and some % info. https://venturebeat.com/technology/claude-codes-source-code-appears-to-have-leaked-heres-what-we-know Internal comments reveal that Anthropic is already iterating on Capybara v8, yet the model still faces significant hurdles. The code notes a 29-30% false claims rate in v8, an actual regression compared to the 16.7% rate seen in v4.
There was some references to experimental models not publicly available and some % info.
https://venturebeat.com/technology/claude-codes-source-code-appears-to-have-leaked-heres-what-we-know