An Industrial-Scale Insurance LLM Achieving Verifiable Domain Mastery and Hallucination Control without Competence Trade-offs
Fuente:
arXiv
Enregistré dans:
| Auteurs principaux: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Publié: |
2026
|
| Sujets: | |
| Accès en ligne: | |
| Tags: |
Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
|
| _version_ | 1866918491942027264 |
|---|---|
| author | Zhu, Qian Guo, Xinnan Huo, Jingjing Li, Jun Liu, Pan Yang, Wenyan Xu, Wanqing Lin, Xuan |
| author_facet | Zhu, Qian Guo, Xinnan Huo, Jingjing Li, Jun Liu, Pan Yang, Wenyan Xu, Wanqing Lin, Xuan |
| contents | Adapting Large Language Models (LLMs) to high-stakes vertical domains like insurance presents a significant challenge: scenarios demand strict adherence to complex regulations and business logic with zero tolerance for hallucinations. Existing approaches often suffer from a Competency Trade-off - sacrificing general intelligence for domain expertise - or rely heavily on RAG without intrinsic reasoning. To bridge this gap, we present INS-S1, an insurance-specific LLM family trained via a novel end-to-end alignment paradigm. Our approach features two methodological innovations: (1) A Verifiable Data Synthesis System that constructs hierarchical datasets for actuarial reasoning and compliance; and (2) A Progressive SFT-RL Curriculum Framework that integrates dynamic data annealing with a synergistic mix of Verified Reasoning (RLVR) and AI Feedback (RLAIF). By optimizing data ratios and reward signals, this framework enforces domain constraints while preventing catastrophic forgetting. Additionally, we release INSEva, the most comprehensive insurance benchmark to date (39k+ samples). Extensive experiments show that INS-S1 achieves SOTA performance on domain tasks, significantly outperforming DeepSeek-R1 and Gemini-2.5-Pro. Crucially, it maintains top-tier general capabilities and achieves a record-low 0.6% hallucination rate (HHEM). Our results demonstrate that rigorous domain specialization can be achieved without compromising general intelligence. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_14463 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | An Industrial-Scale Insurance LLM Achieving Verifiable Domain Mastery and Hallucination Control without Competence Trade-offs Zhu, Qian Guo, Xinnan Huo, Jingjing Li, Jun Liu, Pan Yang, Wenyan Xu, Wanqing Lin, Xuan Computation and Language I.2.7; I.2.1 Adapting Large Language Models (LLMs) to high-stakes vertical domains like insurance presents a significant challenge: scenarios demand strict adherence to complex regulations and business logic with zero tolerance for hallucinations. Existing approaches often suffer from a Competency Trade-off - sacrificing general intelligence for domain expertise - or rely heavily on RAG without intrinsic reasoning. To bridge this gap, we present INS-S1, an insurance-specific LLM family trained via a novel end-to-end alignment paradigm. Our approach features two methodological innovations: (1) A Verifiable Data Synthesis System that constructs hierarchical datasets for actuarial reasoning and compliance; and (2) A Progressive SFT-RL Curriculum Framework that integrates dynamic data annealing with a synergistic mix of Verified Reasoning (RLVR) and AI Feedback (RLAIF). By optimizing data ratios and reward signals, this framework enforces domain constraints while preventing catastrophic forgetting. Additionally, we release INSEva, the most comprehensive insurance benchmark to date (39k+ samples). Extensive experiments show that INS-S1 achieves SOTA performance on domain tasks, significantly outperforming DeepSeek-R1 and Gemini-2.5-Pro. Crucially, it maintains top-tier general capabilities and achieves a record-low 0.6% hallucination rate (HHEM). Our results demonstrate that rigorous domain specialization can be achieved without compromising general intelligence. |
| title | An Industrial-Scale Insurance LLM Achieving Verifiable Domain Mastery and Hallucination Control without Competence Trade-offs |
| topic | Computation and Language I.2.7; I.2.1 |
| url | https://arxiv.org/abs/2603.14463 |