"SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness"
This is a dev post classified by Jev as AI dev tools (a free resource), kept by the Dev Radar because it carries real work, not commentary.
"SoL-Pi: Recursively Scaling Auto-Research Loops for Efficient Agent Harness" This paper lets coding agents recursively optimize their own harness instead of hand-designing it. Across diverse coding environments, it discovers four mechanisms that compress context, observations, reading, and actions. The result is comparable performance with 44.7-49% less token traffic and roughly one-third lower API cost. https://www.alphaxiv.org/abs/2609.20519
Posted by alphaXiv (55.9k followers) 1 h ago · 27 likes · 1.3k views · view the original post on X. Kept by the Dev Radar as AI dev tools. Tools mentioned: alphaXiv.
More dev work like this
- StepFun released Step 5 Preview, a new 600B total / 27B active MoE model with 1M context. — @testingcatalog
- Teaches how to engineer reliable AI coding agents through environment, state management,… — @tom_doerr
- 🚀 Stop generating JSON tokens for classification! We just ran an exhaustive, strictly… — @zast57
- Your coding agents shouldn’t start every session from zero. — @DanKornas
- A 27B AI model running locally inside a browser.🚀 — @BkashJosi
- Introducing a new way FIND (⌘F) with @typesafeai Jev. — @Saboo_Shubham_
- Codex 在家里的 Mac 上跑着一个长任务，人出门了，跑到要审批的那一步就停在那儿等，回来才发现半天没动。 — @GitHub_Daily
- Stepfun just announced their new SOTA model — @notjazii
Every post is read and classified by Jev (TypeSafe): what it is, which market it belongs to, and whether the link is a real tool. 15.3k posts from 4.8k X accounts over the last 21 days, 1.7k tools, 12 markets. Collected every 5 minutes, fully re-ranked every hour — last update 2026-09-20 08:51 UTC. Full method.