Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866912235171872768 |
|---|---|
| author | Ye, Jiacheng Gao, Jiahui Gong, Shansan Zheng, Lin Jiang, Xin Li, Zhenguo Kong, Lingpeng |
| author_facet | Ye, Jiacheng Gao, Jiahui Gong, Shansan Zheng, Lin Jiang, Xin Li, Zhenguo Kong, Lingpeng |
| contents | Autoregressive language models, despite their impressive capabilities, struggle with complex reasoning and long-term planning tasks. We introduce discrete diffusion models as a novel solution to these challenges. Through the lens of subgoal imbalance, we demonstrate how diffusion models effectively learn difficult subgoals that elude autoregressive approaches. We propose Multi-Granularity Diffusion Modeling (MGDM), which prioritizes subgoals based on difficulty during learning. On complex tasks like Countdown, Sudoku, and Boolean Satisfiability Problems, MGDM significantly outperforms autoregressive models without using search techniques. For instance, MGDM achieves 91.5\% and 100\% accuracy on Countdown and Sudoku, respectively, compared to 45.8\% and 20.7\% for autoregressive models. Our work highlights the potential of diffusion-based approaches in advancing AI capabilities for sophisticated language understanding and problem-solving tasks. All associated codes are available at \href{https://github.com/HKUNLP/diffusion-vs-ar}{https://github.com/HKUNLP/diffusion-vs-ar}. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_14157 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning Ye, Jiacheng Gao, Jiahui Gong, Shansan Zheng, Lin Jiang, Xin Li, Zhenguo Kong, Lingpeng Computation and Language Machine Learning Autoregressive language models, despite their impressive capabilities, struggle with complex reasoning and long-term planning tasks. We introduce discrete diffusion models as a novel solution to these challenges. Through the lens of subgoal imbalance, we demonstrate how diffusion models effectively learn difficult subgoals that elude autoregressive approaches. We propose Multi-Granularity Diffusion Modeling (MGDM), which prioritizes subgoals based on difficulty during learning. On complex tasks like Countdown, Sudoku, and Boolean Satisfiability Problems, MGDM significantly outperforms autoregressive models without using search techniques. For instance, MGDM achieves 91.5\% and 100\% accuracy on Countdown and Sudoku, respectively, compared to 45.8\% and 20.7\% for autoregressive models. Our work highlights the potential of diffusion-based approaches in advancing AI capabilities for sophisticated language understanding and problem-solving tasks. All associated codes are available at \href{https://github.com/HKUNLP/diffusion-vs-ar}{https://github.com/HKUNLP/diffusion-vs-ar}. |
| title | Beyond Autoregression: Discrete Diffusion for Complex Reasoning and Planning |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2410.14157 |