Fetching from the wire…
01
02
03
04
05
06
07
08
Research2026-08-28 · source-backed
SWE-Prime's premise is that a successful trajectory still contains ineffective, redundant and risky steps, so SFT on all resolved runs teaches bad habits (arXiv 2608.27449). It filters at trajectory level on process quality, result quality and representativeness, then at segment level by grouping consecutive steps and scoring each on contribution, learnability and risk. All segments stay in the sequence for context, but only selected ones contribute to the loss. The 10% subset beats the full resolved dataset by up to 12.2% on SWE-Bench Pro and 24.2% on Verified.
Each link below shares sources, entities, or timing with this story.
Shared entities / Same source domain / Shared topic / Earlier coverage
Both cover Bench Pro, SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, trajectory).
Both cover SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, beat, level, trajectory).
Shared entities / Same source domain / Shared topic / Earlier coverage / Tension
Both cover Bench Pro, SWE; reported by the same outlet (arxiv.org); overlapping topics (agent, beat).
Shared entities / Same source domain / Shared topic / Earlier coverage
Both cover SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, beat, dataset).
Shared entities / Shared topic / Earlier coverage
Both cover Bench Pro, SWE, Verified; overlapping topics (agent, context); earlier Bench Pro coverage from 2026-03-06.
Shared entities / Same source domain / Shared topic / Earlier coverage
Both cover SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, context).
Both cover SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, context).
Both cover SWE, Verified; reported by the same outlet (arxiv.org); overlapping topics (agent, context).