Skip to content

Formats · Latest news

Papers & research

AI papers and findings worth reading: architecture, training methods, capability measurement, and theory.

161 top picks all-time · 90 in the past 30 days · chosen from 1,274 items collected all-time

Latest pick

Top picks archive · Page 9

Top picks 161–161 of 161

Mar 14, 2024

Mar 14, 2024Thu
  1. Cognition Blog (Devin, Windsurf)OfficialAI score62

    Cognition reports Devin resolves 13.86% of SWE-bench issues end to end

    AICognition reports that its agent Devin resolved 79 of 570 sampled SWE-bench issues, a 13.86% success rate, without being given the files to edit.

    Why it matters: The report explains how SWE-bench was adapted for end-to-end agent evaluation, with failure cases that clarify where the 13.86% result comes from and its limits.