Cognition's SWE-2 achieves 92.8 on Terminal-Bench 2.1

(tokenstead.ai)

24 points | by cdnsteve 2 hours ago

4 comments

  • forgot-my-pw 29 minutes ago
    Terminal Bench 2/2.1 appears to be almost solved, so we probably shouldn't look too hard on that? Their Terminal Bench 4 score is soso.

    Still quite impressive though.

  • ChrisArchitect 3 minutes ago
    Related:

    Cognition launches new SWE-2 model

    https://news.ycombinator.com/item?id=49645443

  • walrus01 11 minutes ago
    Not really news, terminal bench 4 is the new metric. It's only a few points ahead in terminal bench 4 of some open weight models you can run on a 256GB system.
  • varispeed 1 hour ago
    These tests are pointless when often models get nerfed few days after release. Astra today is way dumber than just few days ago.
    • kzrdude 47 minutes ago
      Is that something we have credible evidence for? Do they serve a better model when artificialanalysis (the benchmark site) is making the requests, and so on?
      • varispeed 5 minutes ago
        It is my own anectodal. Few days something I worked on usually got one shotted or got quality result. Today it is very much going nowhere and is stuck in reasoning loops.

        Probably someone should build nerf tracker, because this is quite common that models get substantially worse once PR hype wears off and they quantise them more or simply route requests to older models with system prompt changed to say it is Astra and not Sol etc.

    • cbg0 49 minutes ago
      Not Astra but Sol hasn't been nerfed: https://marginlab.ai/trackers/codex/
      • varispeed 4 minutes ago
        That is not conclusive, because they can detect such tracker and route it through proper not quantised model.
    • d_tr 49 minutes ago
      How and why do they get nerfed? To save money?
      • varispeed 3 minutes ago
        Yes. They save on compute and customer has to use more tokens to achieve their goal which means more profit.