Metraipolis › archive › dc_405b36dce034

TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents — arXiv:2608.22793

Source: arXiv (en)
Original: https://arxiv.org/abs/2608.22793
License (as archived): CC0-1.0
Archived: 2026-10-18T07:45:01.974Z
SHA-256 of the archived text: 945b9a315b2c9521adf1cc4dea105a6f6382ac0cdf9f4d8b07a891a9495545d9
Stable URL of this page: https://metraipolis.com/documento/dc_405b36dce034

How to cite

TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents — arXiv:2608.22793 — arXiv (en), https://arxiv.org/abs/2608.22793 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_405b36dce034 (Metraipolis archive dc_405b36dce034, sha256 945b9a315b2c9521adf1cc4dea105a6f6382ac0cdf9f4d8b07a891a9495545d9)

BibTeX

@misc{dc405b36dce034,
  title = {TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents — arXiv:2608.22793},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2608.22793},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_405b36dce034; sha256 945b9a315b2c9521adf1cc4dea105a6f6382ac0cdf9f4d8b07a891a9495545d9},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.22793. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Reliable deployment of LLM agents in user-facing products depends not on raw task-solving ability but on consistency and limit-awareness: behaving the same way across repeated trials, and recognizing when a request cannot, or cannot yet, be safely fulfilled. CAR-bench exposes this reliability gap in the domain of in-car assistants: an LLM-simulated user issues incomplete or ambiguous requests, requiring the agent to resolve uncertainty through multi-turn dialogue and tool use while strictly adhering to domain policies. Even frontier models show a substantial gap between what they can solve at least once (Pass@3) and what they solve consistently across trials (Pass^k). We bridge this gap with TRACE (TRAjectory-Contrastive Evolution), which iteratively improves a skill-based agent's behavioral knowledge without modifying model weights. This knowledge is organized as a Skill Bank of modular, retrievable skills, each encoding a self-contained set of tool-use rules and behavioral guidelines. TRACE evolves this bank through an agentic self-evolution loop: after each evaluation round, it groups trajectories by the skills invoked and refines each skill by contrasting successful and failed behaviors. The updated bank then guides subsequent rounds, while during deployment the Actor performs state-conditioned skill orchestration at every turn. On GPT-5.5, TRACE improves consistency (Pass^3) by 34.6 points, from 59.9% to 94.5%, while shrinking the gap between potential and reliable performance to just 4.0 points. On the official hidden set, TRACE achieved first place using GPT-5.6-Sol, attaining a Pass^3 score of 70%-a 40% relative improvement over the baseline. These results show that TRACE converts high model potential into stable, consistent performance gain. Project homepage: this https URL .

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-25 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2608.22793 — 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 2026-10-18T07:37:52.899Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-25?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).