Switch-SWE benchmark shows coding agents lose focus when switching unfinished tasks
SJTU · hf · 2026-11-04
SJTU and Tongji introduce Switch-SWE, a repository-level benchmark measuring whether coding agents can resume interrupted tasks. Interleaving unfinished tasks from the same repo with persistent checkouts and executable verifiers, it shows clear performance drops in switched sessions, and a context-management module improves recovery.
More from coding & agent
- Tencent's Hy4 agent self-trained a smaller model that beat Tencent's own baseline — heyshrutimishra · 2026-11-04
- Butterfly traces multi-agent failures with graphs, 62.5% accuracy vs 53.8% for log prompting — JeremyCMorgan · 2026-11-04
- NatWest on agent safety: an 83% human-review pass rate hid a 54-point gap — AI Engineer · 2026-11-04
- Free course covers the 4 hard problems of production generative media agents — Al_Grigor · 2026-11-04
- Quota reset night: Codex back to 100% while Claude shows just 10% left — rudrank · 2026-11-04
- Self-improving software is inevitable: closing the agent-observability loop — bibryam · 2026-11-04