Stories

Completed digest stories linked to this service.

• SWE-bench scores are spiking, but variant mix-ups make the leaderboard noisy for...

  2026-04-12

  Vendors are touting big SWE-bench jumps, but versions differ and scores alone won’t pick your coding copilot. ...
• Anthropic’s Mythos and Project Glasswing push AI into real-world vuln discovery,...

  2026-04-09

  Anthropic launched Project Glasswing and a Mythos Preview model that finds serious software bugs, pairing indu...
• Claude Mythos posts record SWE-bench numbers, but it’s gated; tighten your evals...

  2026-04-08

  Anthropic’s Claude Mythos preview claims record SWE-bench results, but it isn’t publicly available and public ...
• SWE-Bench Pro leaderboard: small gains at the top, big contexts, and mostly self...

  2026-04-04

  A new SWE-Bench Pro leaderboard shows top code models clustered around 0.55–0.58, with large contexts and self...
• Coding-agent benchmarks are wobbling—trust results only after your own cross-con...

  2026-03-24

  SWE-Bench-style coding scores are spiking, but contamination and self-reported leaderboards mean you should tr...