deepswe

Latest stories covering deepswe, with summaries and links to related coverage.

2 articles · page 1 of 1

• Laguna S 2.1

  2026-07-21

  tool

  19 min read

  Laguna S 2.1 has been released, enhancing model capabilities for longer horizon tasks. The update includes improvements in training methodology and post-training task distribution.

  lagunaai-modelspost-training

• DeepSWE: A contamination-free benchmark for long-horizon coding agents

  2026-05-26

  research

  20 min read

  DeepSWE is a long-horizon software engineering benchmark designed to evaluate coding agents on original engineering tasks. It features contamination-free tasks, high diversity across 91 repositories in five programming languages, and real-world applicability.

  software-engineeringai-agentscode-generation