FrogNano: a 4B model trained purely with RL on synthetic tasks hits repo-level coding
burkov · x · 2026-09-10
FrogNano shows a compact 4B model can reach strong repository-level coding performance via RL on synthetic tasks alone, without distillation from larger models. The pipeline pairs TaskPilot, an online task-synthesis system that generates executable SWE tasks from real repo snapshots calibrated to the policy's learnability frontier, with Leaf, a lightweight harness using five typed tools and a simple termination rule. Base Qwen3.5-4B is trained with group-relative DPPO over five successive rounds of task synthesis.
More from coding & agent
- Firecrawl offers 2.5x credits to a16z Speedrun founders, unveils $75M Series B — devdigest · 2026-10-27
- Chums: open-source Apple Watch app puts Claude Code and Codex on your wrist — brandon_galang · 2026-10-27
- Open-source Clairic Studio turns pasted text into typed data and working code in 5 languages — dfinke · 2026-10-27
- Matt Pocock shares expand/contract trick: split one-way-door PRs to de-risk big changes — mattpocockuk · 2026-10-27
- Qdrant builds Universal I/O to run vector search mmap code on object storage — qdrant_engine · 2026-10-27
- theo publishes tstack skills, Garry Tan urges everyone to 'publish your stack' — ycombinator · 2026-10-26