Salvato in:
| Autore principale: | |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | https://arxiv.org/abs/2510.16309 |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914197868118016 |
|---|---|
| author | Su, Crystal |
| author_facet | Su, Crystal |
| contents | Large language models (LLMs) often produce fluent reasoning steps while violating simple mathematical or logical constraints. We introduce MedRule-KG, a compact typed knowledge graph coupled with a symbolic verifier, designed to enforce mathematically interpretable rules in reasoning tasks. MedRule-KG encodes entities, relations, and three domain-inspired rules, while the verifier checks predictions and applies minimal corrections to guarantee consistency. On a 90-example FDA-derived benchmark, grounding in MedRule-KG improves exact match (EM) from 0.767 to 0.900, and adding the verifier yields 1.000 EM while eliminating rule violations entirely. We demonstrate how MedRule-KG provides a general scaffold for safe mathematical reasoning, discuss ablations, and release code and data to encourage reproducibility. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2510_16309 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | MedRule-KG: A Knowledge-Graph--Steered Scaffold for Mathematical Reasoning with a Lightweight Verifier Su, Crystal Artificial Intelligence Large language models (LLMs) often produce fluent reasoning steps while violating simple mathematical or logical constraints. We introduce MedRule-KG, a compact typed knowledge graph coupled with a symbolic verifier, designed to enforce mathematically interpretable rules in reasoning tasks. MedRule-KG encodes entities, relations, and three domain-inspired rules, while the verifier checks predictions and applies minimal corrections to guarantee consistency. On a 90-example FDA-derived benchmark, grounding in MedRule-KG improves exact match (EM) from 0.767 to 0.900, and adding the verifier yields 1.000 EM while eliminating rule violations entirely. We demonstrate how MedRule-KG provides a general scaffold for safe mathematical reasoning, discuss ablations, and release code and data to encourage reproducibility. |
| title | MedRule-KG: A Knowledge-Graph--Steered Scaffold for Mathematical Reasoning with a Lightweight Verifier |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2510.16309 |