Metraipolis › archive › dc_9d85d2ef54e8

Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards — arXiv:2610.22448

Source
arXiv (en)
Original
https://arxiv.org/abs/2610.22448
License (as archived)
CC0-1.0
Archived
2026-10-19T08:48:18.024Z
SHA-256 of the archived text
8715230c6eeb484b46523b60044ff5ac8f00ec8869ff3b7faf5f9d9747501a45
Stable URL of this page
https://metraipolis.com/documento/dc_9d85d2ef54e8

How to cite

Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards — arXiv:2610.22448 — arXiv (en), https://arxiv.org/abs/2610.22448 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_9d85d2ef54e8 (Metraipolis archive dc_9d85d2ef54e8, sha256 8715230c6eeb484b46523b60044ff5ac8f00ec8869ff3b7faf5f9d9747501a45)
BibTeX
@misc{dc9d85d2ef54e8,
  title = {Lean4AgenticRL: Advancing Agentic RL with Verifiable Multi-level Process Rewards — arXiv:2610.22448},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2610.22448},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_9d85d2ef54e8; sha256 8715230c6eeb484b46523b60044ff5ac8f00ec8869ff3b7faf5f9d9747501a45},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.22448. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Enhancing the agentic capabilities of Large Language Models (LLMs) to tackle complex software engineering (SWE) and general terminal-usage problems is a pressing challenge. Current state-of-the-art LLMs often go through large-scale agentic Reinforcement Learning (RL). However, they mostly rely on high-level outcome rewards, with minimal consideration of low-level toolcall and mid-level workflow-stage rewards, which makes the training less effective. To overcome this limitation, we propose **Lean4AgenticRL**, which, to the best of our knowledge, is the first framework that grounds multi-level process rewards in the unified Lean4 formalization schema for open-ended agentic RL. **Lean4AgenticRL** first builds a *problem-specific workflow* system that balances the agent's own capability and controllability during execution, laying the foundation for further verification. Subsequently, our workflow formalization and multi-level reward annotation methods enable agent behavior verification within a general harness system. It provides multi-level rewards to support more effective agentic RL with the workflow-stage-normalized GRPO algorithm. Using **Lean4AgenticRL**, we train **LAR-9B/27B**, a family of models with enhanced agentic capability. Extensive experiments indicate that our model outperforms the leading model of similar size. **LAR-27B** achieves **77.3%** on TerminalBench-2.1. On average, our models outperform the base model by **9.89%** and surpass the second-best model of comparable size we know by **5.07%**. These results indicate the potential of integrating formal methods in enhancing open-ended agentic RL.

Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno).
Categorie arXiv: Machine Learning (cs.LG).
Pagina del lavoro: https://arxiv.org/abs/2610.22448 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito).
Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina.
Letto il 2026-10-19T08:42:40.206Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).