Test-time Scaling of LLMs: A Survey from A Subproblem Structure Perspective
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866917090122792960 |
|---|---|
| author | Yang, Zhuoyi Guo, Xu Zhang, Tong Xu, Huijuan Li, Boyang |
| author_facet | Yang, Zhuoyi Guo, Xu Zhang, Tong Xu, Huijuan Li, Boyang |
| contents | With this paper, we survey techniques for improving the predictive accuracy of pretrained large language models by allocating additional compute at inference time. In categorizing test-time scaling methods, we place special emphasis on how a problem is decomposed into subproblems and on the topological organization of these subproblems whether sequential, parallel, or tree-structured. This perspective allows us to unify diverse approaches such as Chain-of-Thought, Branch-Solve-Merge, and Tree-of-Thought under a common lens. We further synthesize existing analyses of these techniques, highlighting their respective strengths and weaknesses, and conclude by outlining promising directions for future research |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2511_14772 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Test-time Scaling of LLMs: A Survey from A Subproblem Structure Perspective Yang, Zhuoyi Guo, Xu Zhang, Tong Xu, Huijuan Li, Boyang Computation and Language Artificial Intelligence With this paper, we survey techniques for improving the predictive accuracy of pretrained large language models by allocating additional compute at inference time. In categorizing test-time scaling methods, we place special emphasis on how a problem is decomposed into subproblems and on the topological organization of these subproblems whether sequential, parallel, or tree-structured. This perspective allows us to unify diverse approaches such as Chain-of-Thought, Branch-Solve-Merge, and Tree-of-Thought under a common lens. We further synthesize existing analyses of these techniques, highlighting their respective strengths and weaknesses, and conclude by outlining promising directions for future research |
| title | Test-time Scaling of LLMs: A Survey from A Subproblem Structure Perspective |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2511.14772 |