Ai assistants can fix the bugs they cannot find. a new swe-explore benchmark, out september 17,…
ai assistants can fix the bugs they cannot find. a new swe-explore benchmark, out september 17, shows coding assistants fall to 14-19% accuracy at line-level bug localization even when they repair whole files well.
finding the bug is now the harder half.
Context
The SWE-Explore preprint (arXiv 2606.07297) builds a benchmark of 848 issues across 10 languages and 203 repositories that isolates repository exploration: explorers return ranked code regions under a line budget, with ground truth derived from independent agent trajectories that solved each issue. The paper text says general-purpose coding agents all reach high HitFile and nDCG@500 while their line-level recall (Rec-l) stays around 0.14 to 0.19.
The 0.14 to 0.19 range is line-level recall, not accuracy, so accuracy is the note's wording. Out September 17 is not supported: search metadata dates the paper 5 June 2026 and the repository 8 June 2026. Repair whole files well rests on strong file-level hit rates in the paper, and no measured whole-file repair result was confirmed. Results use a Mini-SWE-Agent scaffold across models. Preprint, not peer reviewed.
Watch next
- A dated release for the benchmark and any whole-file repair measurement.
Sources
- SWE-Explore (arXiv 2606.07297)arxiv.org
Provenance
The note above is reproduced unedited from the original post, first published on Threads on 20 September 2026 at 20:16 IST. Sources are the papers and datasets the note draws on.
View the original post ↗Embed this note
More notes
The air is now being asked to keep its own ledger
the air is now being asked to keep its own ledger: ecmwf’s aifs compo becomes the first ai model to forecast atmospheric composition globally every three hours, cleanair simulates 365 days of pm2.5 over china in ten seconds, and a unified framework maps six pollutants at one kilometer across the whole country. the air now files its own composition report.
read the note →The current is now being asked to draw its own map
the current is now being asked to draw its own map: china’s langya 2.0 predicts six ocean phenomena including internal waves and mesoscale eddies, a deep net called wenhai resolves eddies globally with air sea flux formulas built in, and scripps infers surface currents from the way temperature patterns deform in satellite images. the ocean now files its own circulation report.
read the note →The soil is now being asked to report its own carbon
the soil is now being asked to report its own carbon: a nix color sensor paired with generative data augmentation predicts soil organic carbon without a lab, random forest drives 74 percent of soil health mapping studies, and sentinel 2 tracks five year carbon change across france and italy from 922 samples. the dirt now files its own carbon account.
read the note →