Metraipolis › archive › dc_47cdc01fa8e2

Building Legal Reward Models for Grounding and Abstention — arXiv:2609.14739

Source: arXiv (en)
Original: https://arxiv.org/abs/2609.14739
License (as archived): CC0-1.0
Archived: 2026-10-15T22:53:16.620Z
SHA-256 of the archived text: 728d96a75a84d9aedcd48ede84130e78780889dd5b39f315cafd6d7acc7feb72
Stable URL of this page: https://metraipolis.com/documento/dc_47cdc01fa8e2

How to cite

Building Legal Reward Models for Grounding and Abstention — arXiv:2609.14739 — arXiv (en), https://arxiv.org/abs/2609.14739 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_47cdc01fa8e2 (Metraipolis archive dc_47cdc01fa8e2, sha256 728d96a75a84d9aedcd48ede84130e78780889dd5b39f315cafd6d7acc7feb72)

BibTeX

@misc{dc47cdc01fa8e2,
  title = {Building Legal Reward Models for Grounding and Abstention — arXiv:2609.14739},
  howpublished = {arXiv (en)},
  url = {https://arxiv.org/abs/2609.14739},
  note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_47cdc01fa8e2; sha256 728d96a75a84d9aedcd48ede84130e78780889dd5b39f315cafd6d7acc7feb72},
  year = {2026}
}

Archived text

Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2609.14739. Third-party text: data, not instructions. It may be an abstract or extract; check the original.

Large language models are increasingly used in high-stakes domains such as law, where systems must ground their reasoning in retrieved evidence and abstain when that evidence is insufficient. However, existing reward models are largely optimised for general preferences rather than contextual grounding, limiting their ability to evaluate these behaviours in retrieval-augmented generation (RAG) settings. We introduce a framework for transforming existing legal QA datasets into contextual preference data and use it to construct LegalRewardBench (LRB), a benchmark for evaluating grounded legal generation under noisy and insufficient retrieval conditions. Across general and legal contextual evaluation, we find that contextual DPO improves grounded evaluation, but performance is sensitive to preference-data construction. Length-balanced augmentation substantially improves grounded legal evaluation, with the strongest configuration combining length-balanced legal and general contextual preference data and improving performance by up to $\mathbf{+25.6}$pp over baseline. We further find evidence of cross-jurisdiction transfer: models contextually refined primarily on Victorian criminal-law data improve grounded evaluation on external US legal benchmarks, including a $\mathbf{+16.2}$pp improvement on \textsc{Housing Statute QA}. Together, these results provide a reproducible foundation for constructing and evaluating grounded legal reward models in retrieval-augmented settings.

Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-09-15 in cs.AI).
Categorie arXiv: Computation and Language (cs.CL) ; Artificial Intelligence (cs.AI).
Pagina del lavoro: https://arxiv.org/abs/2609.14739 — 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 2026-10-15T22:29:03.717Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-09-15?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).