5

Source document

[2026-02-18 23:21] ajaxdavis: of course [2026-02-18 23:21] ajaxdavis: https://evals.blah.dev/runs/P7o0Sk-gKOnK75Dsb81LjA [2026-02-18 23:21] ajaxdavis: it is [2026-02-18 23:21] ajaxdavis: omega is not even a model and it's competing lol [2026-02-18 23:21] ajaxdavis: (files: Screenshot_2026-02-19_at_9.21.38_AM.png) [2026-02-18 23:22] foxhop.: im lost [2026-02-18 23:22] foxhop.: (files: Screenshot_from_2026-02-18_18-22-04.png) [2026-02-18 23:22] ajaxdavis: now go to evals.blah.dev [2026-02-18 23:22] ajaxdavis: and oauth with your blah account [2026-02-18 23:22] ajaxdavis: that form is if you want to add blah oauth login to your apps [2026-02-18 23:24] foxhop.: https://evals.blah.dev/models/d6DQsKXVxDNWMpDpRa4l1g [2026-02-18 23:25] ajaxdavis: now trigger a new run [2026-02-18 23:25] ajaxdavis: did you test the connection before creating the model too? [2026-02-18 23:25] foxhop.: yeah [2026-02-18 23:25] ajaxdavis: saweet [2026-02-18 23:25] ajaxdavis: trigger a run [2026-02-18 23:26] foxhop.: https://evals.blah.dev/runs/rRieHaWUApafePJTiwRGTA [2026-02-18 23:27] foxhop.: we are rich [2026-02-18 23:27] ajaxdavis: yeah i was already looking at it populating [2026-02-18 23:30] ajaxdavis: will remove the bad evals and models, and going to do a ux overhaul. [2026-02-18 23:30] foxhop.: ok ask claude about vision tests [2026-02-18 23:30] foxhop.: and music tests [2026-02-18 23:31] foxhop.: https://evals.blah.dev/models/kBGN3lqyoBV_0Iwn7U02sA [2026-02-18 23:36] ajaxdavis: adding to google docs [2026-02-18 23:58] ajaxdavis: add all your ideas for eval arena here https://docs.google.com/document/d/1uazLfNE0vfbAf4A0pPFBv6ai2fJTy__0FAh05cuehRY/edit?tab=t.0 [2026-02-18 23:59] foxhop.: bless my edits [2026-02-19 00:03] ajaxdavis: anyone can edit now, refresh [2026-02-19 00:51] traves_theberge: it doesnt look that bad! [2026-02-19 01:35] ajaxdavis: (files: Screenshot_from_2026-02-19_11-35-16.png) [2026-02-19 01:39] ajaxdavis: https://alpha.omegaai.dev/runs/novels-run