Florian Brand@xeophon4hOn the public TerminalBench traces, Astra runs into safety classifiers and their rollouts get stopped and scored as 0 Fable gets re-routed to Opus instead and the run continues. Hard to compare scores that way!43482K
Florian Brand@xeophon15hi like Gemini for this use case over a lot of models, incl. Muse Spark guess its now the Claudish translate model, oh wellFlorian Brand@xeophonSep 13asking Gemini to translate swarm<->english is surprisingly effective21704.8K
Florian Brand@xeophon20hbig sandbox doesn't want you to know this but you can bake dependencies into the images you don't have to use python:3.11-slim and download at runtime26830316K
Florian Brand@xeophon23hbecause agents are so good at managing other agents, the best form factor for coding aren't a bunch of projects but just a chat box (again)2703.5K
Florian Brand@xeophonSep 14why do i, the evals guy, have to be dragged into this and now have to debug the parsing 😔Florian Brand@xeophonSep 10i showed the inference guys the new arch and they blocked me and reported me to hr for harassment4815.2K
Florian Brand@xeophonSep 10i showed the inference guys the new arch and they blocked me and reported me to hr for harassment
