UnMaskFork: Test-Time Scaling for Masked Diffusion via Deterministic Action Branching
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_ | 1866912875555061760 |
|---|---|
| author | Misaki, Kou Akiba, Takuya |
| author_facet | Misaki, Kou Akiba, Takuya |
| contents | Test-time scaling strategies have effectively leveraged inference-time compute to enhance the reasoning abilities of Autoregressive Large Language Models. In this work, we demonstrate that Masked Diffusion Language Models (MDLMs) are inherently amenable to advanced search strategies, owing to their iterative and non-autoregressive generation process. To leverage this, we propose UnMaskFork (UMF), a framework that formulates the unmasking trajectory as a search tree and employs Monte Carlo Tree Search to optimize the generation path. In contrast to standard scaling methods relying on stochastic sampling, UMF explores the search space through deterministic partial unmasking actions performed by multiple MDLMs. Our empirical evaluation demonstrates that UMF consistently outperforms existing test-time scaling baselines on complex coding benchmarks, while also exhibiting strong scalability on mathematical reasoning tasks. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_04344 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | UnMaskFork: Test-Time Scaling for Masked Diffusion via Deterministic Action Branching Misaki, Kou Akiba, Takuya Machine Learning Artificial Intelligence Test-time scaling strategies have effectively leveraged inference-time compute to enhance the reasoning abilities of Autoregressive Large Language Models. In this work, we demonstrate that Masked Diffusion Language Models (MDLMs) are inherently amenable to advanced search strategies, owing to their iterative and non-autoregressive generation process. To leverage this, we propose UnMaskFork (UMF), a framework that formulates the unmasking trajectory as a search tree and employs Monte Carlo Tree Search to optimize the generation path. In contrast to standard scaling methods relying on stochastic sampling, UMF explores the search space through deterministic partial unmasking actions performed by multiple MDLMs. Our empirical evaluation demonstrates that UMF consistently outperforms existing test-time scaling baselines on complex coding benchmarks, while also exhibiting strong scalability on mathematical reasoning tasks. |
| title | UnMaskFork: Test-Time Scaling for Masked Diffusion via Deterministic Action Branching |
| topic | Machine Learning Artificial Intelligence |
| url | https://arxiv.org/abs/2602.04344 |