All Roads Lead to Rome: Graph-Based Confidence Estimation for Large Language Model Reasoning
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2025
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866914039900143616 |
|---|---|
| author | Zhang, Caiqi Shu, Chang Shareghi, Ehsan Collier, Nigel |
| author_facet | Zhang, Caiqi Shu, Chang Shareghi, Ehsan Collier, Nigel |
| contents | Confidence estimation is essential for the reliable deployment of large language models (LLMs). Existing methods are primarily designed for factual QA tasks and often fail to generalize to reasoning tasks. To address this gap, we propose a set of training-free, graph-based confidence estimation methods tailored to reasoning tasks. Our approach models reasoning paths as directed graphs and estimates confidence by exploiting graph properties such as centrality, path convergence, and path weighting. Experiments with two LLMs on three reasoning datasets demonstrate improved confidence estimation and enhanced performance on two downstream tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_12908 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | All Roads Lead to Rome: Graph-Based Confidence Estimation for Large Language Model Reasoning Zhang, Caiqi Shu, Chang Shareghi, Ehsan Collier, Nigel Computation and Language Artificial Intelligence Confidence estimation is essential for the reliable deployment of large language models (LLMs). Existing methods are primarily designed for factual QA tasks and often fail to generalize to reasoning tasks. To address this gap, we propose a set of training-free, graph-based confidence estimation methods tailored to reasoning tasks. Our approach models reasoning paths as directed graphs and estimates confidence by exploiting graph properties such as centrality, path convergence, and path weighting. Experiments with two LLMs on three reasoning datasets demonstrate improved confidence estimation and enhanced performance on two downstream tasks. |
| title | All Roads Lead to Rome: Graph-Based Confidence Estimation for Large Language Model Reasoning |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2509.12908 |