Metraipolis › archive › dc_37e1bf33fcfd

TRCA: Transition-wise Rubric Credit Assignment for Long-horizon LLM Agents — arXiv:2608.16156

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.16156
License (as archived): CC0-1.0
Archived: 2026-10-18T17:42:07.279Z
SHA-256 of the archived text: 9ae8a49456444cfb2d433f5896bca464b15d19c77af254b5d984efbc64095a54
Stable URL of this page: https://metraipolis.com/documento/dc_37e1bf33fcfd

How to cite

TRCA: Transition-wise Rubric Credit Assignment for Long-horizon LLM Agents — arXiv:2608.16156 — arXiv (en), https://arxiv.org/abs/2608.16156 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_37e1bf33fcfd (Metraipolis archive dc_37e1bf33fcfd, sha256 9ae8a49456444cfb2d433f5896bca464b15d19c77af254b5d984efbc64095a54)

BibTeX

@misc{dc37e1bf33fcfd,
  title = {TRCA: Transition-wise Rubric Credit Assignment for Long-horizon LLM Agents — arXiv:2608.16156},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.16156},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_37e1bf33fcfd; sha256 9ae8a49456444cfb2d433f5896bca464b15d19c77af254b5d984efbc64095a54},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.16156. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Long-horizon large language model (LLM) agents are typically optimized with sparse terminal outcomes, making fine-grained credit assignment across multi-step interactions difficult. Existing approaches either rely on process evaluators, which incur annotation and inference costs, or derive step-level credit from successful trajectories. However, successful trajectories are extremely scarce during early-stage reinforcement learning, substantially weakening anchor-based methods. We propose Transition-wise Rubric Credit Assignment (TRCA), which derives step-level supervision directly from action-induced transitions without learned evaluators or successful anchors. TRCA evaluates each transition using Evidence, Execution, and Invalidity rubrics to capture task-relevant information acquisition, valid task execution, and invalid or regressive behavior. From these judgments, Foundational Rubric Reward measures local transition quality, while Breakthrough Rubric Reward tracks newly covered Evidence and Execution conditions to reward incremental task progress. Combined with terminal outcomes, these signals produce fine-grained step-level advantages for policy optimization. Experiments on ALFWorld, WebShop, and seven search-augmented question-answering benchmarks show consistent improvements over the evaluated baselines. With Qwen2.5-7B-Instruct, TRCA improves the WebShop score by 6.0%-12.6%; with Qwen2.5-3B-Instruct, it improves the average SearchQA score by 1.9%-18.3%. These results demonstrate the effectiveness of transition-wise rubric credit assignment for long-horizon tasks with sparse successful anchors.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-18 in cs.AI).
Categorie arXiv: Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.16156 — 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 2026-10-18T16:29:13.856Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-18?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).