Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2403.13312 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910375064109056 |
|---|---|
| author | Jiang, Dongwei Fonseca, Marcio Cohen, Shay B. |
| author_facet | Jiang, Dongwei Fonseca, Marcio Cohen, Shay B. |
| contents | Large language models (LLMs) often struggle with complex logical reasoning due to logical inconsistencies and the inherent difficulty of such reasoning. We use Lean, a theorem proving framework, to address these challenges. By formalizing logical reasoning problems into theorems within Lean, we can solve them by proving or disproving the corresponding theorems. This method reduces the risk of logical inconsistencies with the help of Lean's symbolic solver. It also enhances our ability to treat complex reasoning tasks by using Lean's extensive library of theorem proofs. Our method achieves state-of-the-art performance on the FOLIO dataset and achieves performance near this level on ProofWriter. Notably, these results were accomplished by fine-tuning on fewer than 100 in-domain samples for each dataset. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_13312 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | LeanReasoner: Boosting Complex Logical Reasoning with Lean Jiang, Dongwei Fonseca, Marcio Cohen, Shay B. Computation and Language Large language models (LLMs) often struggle with complex logical reasoning due to logical inconsistencies and the inherent difficulty of such reasoning. We use Lean, a theorem proving framework, to address these challenges. By formalizing logical reasoning problems into theorems within Lean, we can solve them by proving or disproving the corresponding theorems. This method reduces the risk of logical inconsistencies with the help of Lean's symbolic solver. It also enhances our ability to treat complex reasoning tasks by using Lean's extensive library of theorem proofs. Our method achieves state-of-the-art performance on the FOLIO dataset and achieves performance near this level on ProofWriter. Notably, these results were accomplished by fine-tuning on fewer than 100 in-domain samples for each dataset. |
| title | LeanReasoner: Boosting Complex Logical Reasoning with Lean |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2403.13312 |