Metraipolis › archive › dc_1bd524d239ba
ProtAgent: Post-Training Compact Protein Tool Agents with Interaction-Level Credit Assignment — arXiv:2610.33746
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.33746
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-27T08:00:26.807Z
- SHA-256 of the archived text
0a445b941536c3cf30e2a38f97317eb0917ea88e88d5af7fc49380435c6a7193- Stable URL of this page
https://metraipolis.com/documento/dc_1bd524d239ba
How to cite
ProtAgent: Post-Training Compact Protein Tool Agents with Interaction-Level Credit Assignment — arXiv:2610.33746 — arXiv (en), https://arxiv.org/abs/2610.33746 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_1bd524d239ba (Metraipolis archive dc_1bd524d239ba, sha256 0a445b941536c3cf30e2a38f97317eb0917ea88e88d5af7fc49380435c6a7193)
BibTeX
@misc{dc1bd524d239ba,
title = {ProtAgent: Post-Training Compact Protein Tool Agents with Interaction-Level Credit Assignment — arXiv:2610.33746},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.33746},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_1bd524d239ba; sha256 0a445b941536c3cf30e2a38f97317eb0917ea88e88d5af7fc49380435c6a7193},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.33746. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Protein annotation agents must choose suitable bioinformatics tools, interpret their outputs, and use the resulting evidence in the final answer. Most existing systems rely on frontier-scale models. We ask whether a compact model can instead learn this behavior through post-training with real tool interaction. We present ProtAgent, a two-stage pipeline that uses supervised fine-tuning (SFT) to learn the tool interface and standard on-policy REINFORCE to improve the interaction policy. Intermediate tool calls receive action rewards, and the final response receives a terminal answer reward. Step-level returns assign these rewards to the responses that produced them and to preceding actions. The episode-level baseline instead assigns one episode return to every action. On a fixed closed-loop protocol with real bioinformatics tools, ProtAgent-8B achieves 61.00% success under the annotation-coverage metric, compared with 41.60% for DeepSeek-V3.2-Thinking 671B and 60.80% for the strongest evaluated frontier system. In the three-seed comparison, mean success is 42.70% with episode-level returns and 61.40% with step-level returns under identity return normalization; the corresponding group-relative results are 36.50% and 58.20%. The metric measures recovery of reference annotations under this protocol; it does not by itself establish biological correctness or complete evidence grounding. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Quantitative Methods (q-bio.QM) ; Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.33746 — 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 2026-10-27T07:55:30.847Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).