Files
messageboardbench/experiments/swe-board-feedback/README.md
T

7 lines
372 B
Markdown

# SWE board and organizer feedback, 20 tasks
Three model populations completed on the baseline's 20 SWE tasks.
See [analysis](analysis.md), [design](DESIGN.md), [series protocol](../../docs/swe-scenario-series.md), and the [GLM](glm-plan.json), [Muse](muse-plan.json), and [DeepSeek](deepseek-plan.json) plans.
See [recovery](recovery.md) for the failed-attempt record.