Metraipolis › archive › dc_45ebf487a7f2
CIRCE Divines LLM Task Complexity from Constraint Structure — arXiv:2610.41089
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2610.41089
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-02T07:28:49.038Z
- SHA-256 of the archived text
1bbc67fa8c7e77f47b425a9938e96d83dcc326e94feff977053e05d0681ed864- Stable URL of this page
https://metraipolis.com/documento/dc_45ebf487a7f2
How to cite
CIRCE Divines LLM Task Complexity from Constraint Structure — arXiv:2610.41089 — arXiv (en), https://arxiv.org/abs/2610.41089 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_45ebf487a7f2 (Metraipolis archive dc_45ebf487a7f2, sha256 1bbc67fa8c7e77f47b425a9938e96d83dcc326e94feff977053e05d0681ed864)
BibTeX
@misc{dc45ebf487a7f2,
title = {CIRCE Divines LLM Task Complexity from Constraint Structure — arXiv:2610.41089},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2610.41089},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_45ebf487a7f2; sha256 1bbc67fa8c7e77f47b425a9938e96d83dcc326e94feff977053e05d0681ed864},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2610.41089. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Estimating whether a problem is too complex for a large language model (LLM) before the model attempts it can guide model routing, how to allocate test-time compute, and suggest the need for new tools to aid the agent. While existing estimators rely on the problem text, LLM-annotated task difficulty labels, or signals obtained after a model attempts to solve the problem, such as its hidden states, none explicitly leverage the constraint structure inherent in the problem: the variables to assign and how the constraints over them interact. We introduce CIRCE, which extracts a problem's structure by using a cheap LLM to construct a Constraint Intermediate Representation that deterministically compiles into the constraint graph. Across seven reasoning benchmarks (including SAT, logic, and math) and four LLMs, even a randomly initialized GNN encoder improves predictions of LLM response correctness, reducing negative log-likelihood from 0.465 to 0.423 and increasing AUROC from 0.691 to 0.727 relative to the strongest baseline for each metric. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.AI, nuovi lavori del giorno). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2610.41089 — 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 2026-11-02T07:26:57.670Z dalla pagina https://arxiv.org/list/cs.AI/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).