Training Data & Post-TrainingSiwei Wu et al.
Researchers release ModularRSI
ModularRSI released — splits an agent harness into five functional modules and lets each evolve independently within a limited scope of modification. It builds a separate set of 2,000 evolution tasks that do not overlap with the evaluation benchmarks, separating "benchmark-specific adaptation" from "reusable improvement", and shows on TB2.0 and SWE-Bench Verified that the gains transfer to unseen domains and to different base models.
Why it matters
A head-on attack, through data separation, on the chronic problem of harness self-improvement: "fitting only the benchmark". Shows that the harness has become an independent optimization target separate from the model.
Source — Primary sources first; secondary reports are labelled “Press”
Paperarxiv.orghttps://arxiv.org/abs/2609.14857Added in Week 39, 2026 · See that week