From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models
Fuente:
arXiv
Guardado en:
| Autores principales: | , , , , , , , |
|---|---|
| Formato: | Preprint |
| Publicado: |
2024
|
| Materias: | |
| Acceso en línea: | |
| Etiquetas: |
Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
|
| _version_ | 1866929598698094592 |
|---|---|
| author | Welleck, Sean Bertsch, Amanda Finlayson, Matthew Schoelkopf, Hailey Xie, Alex Neubig, Graham Kulikov, Ilia Harchaoui, Zaid |
| author_facet | Welleck, Sean Bertsch, Amanda Finlayson, Matthew Schoelkopf, Hailey Xie, Alex Neubig, Graham Kulikov, Ilia Harchaoui, Zaid |
| contents | One of the most striking findings in modern research on large language models (LLMs) is that scaling up compute during training leads to better results. However, less attention has been given to the benefits of scaling compute during inference. This survey focuses on these inference-time approaches. We explore three areas under a unified mathematical formalism: token-level generation algorithms, meta-generation algorithms, and efficient generation. Token-level generation algorithms, often called decoding algorithms, operate by sampling a single token at a time or constructing a token-level search space and then selecting an output. These methods typically assume access to a language model's logits, next-token distributions, or probability scores. Meta-generation algorithms work on partial or full sequences, incorporating domain knowledge, enabling backtracking, and integrating external information. Efficient generation methods aim to reduce token costs and improve the speed of generation. Our survey unifies perspectives from three research communities: traditional natural language processing, modern LLMs, and machine learning systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_16838 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models Welleck, Sean Bertsch, Amanda Finlayson, Matthew Schoelkopf, Hailey Xie, Alex Neubig, Graham Kulikov, Ilia Harchaoui, Zaid Computation and Language Machine Learning One of the most striking findings in modern research on large language models (LLMs) is that scaling up compute during training leads to better results. However, less attention has been given to the benefits of scaling compute during inference. This survey focuses on these inference-time approaches. We explore three areas under a unified mathematical formalism: token-level generation algorithms, meta-generation algorithms, and efficient generation. Token-level generation algorithms, often called decoding algorithms, operate by sampling a single token at a time or constructing a token-level search space and then selecting an output. These methods typically assume access to a language model's logits, next-token distributions, or probability scores. Meta-generation algorithms work on partial or full sequences, incorporating domain knowledge, enabling backtracking, and integrating external information. Efficient generation methods aim to reduce token costs and improve the speed of generation. Our survey unifies perspectives from three research communities: traditional natural language processing, modern LLMs, and machine learning systems. |
| title | From Decoding to Meta-Generation: Inference-time Algorithms for Large Language Models |
| topic | Computation and Language Machine Learning |
| url | https://arxiv.org/abs/2406.16838 |