Small But Funny: A Feedback-Driven Approach to Humor Distillation
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911785403023360 |
|---|---|
| author | Ravi, Sahithya Huber, Patrick Shrivastava, Akshat Sagar, Aditya Aly, Ahmed Shwartz, Vered Einolghozati, Arash |
| author_facet | Ravi, Sahithya Huber, Patrick Shrivastava, Akshat Sagar, Aditya Aly, Ahmed Shwartz, Vered Einolghozati, Arash |
| contents | The emergence of Large Language Models (LLMs) has brought to light promising language generation capabilities, particularly in performing tasks like complex reasoning and creative writing. Consequently, distillation through imitation of teacher responses has emerged as a popular technique to transfer knowledge from LLMs to more accessible, Small Language Models (SLMs). While this works well for simpler tasks, there is a substantial performance gap on tasks requiring intricate language comprehension and creativity, such as humor generation. We hypothesize that this gap may stem from the fact that creative tasks might be hard to learn by imitation alone and explore whether an approach, involving supplementary guidance from the teacher, could yield higher performance. To address this, we study the effect of assigning a dual role to the LLM - as a "teacher" generating data, as well as a "critic" evaluating the student's performance. Our experiments on humor generation reveal that the incorporation of feedback significantly narrows the performance gap between SLMs and their larger counterparts compared to merely relying on imitation. As a result, our research highlights the potential of using feedback as an additional dimension to data when transferring complex language abilities via distillation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2402_18113 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Small But Funny: A Feedback-Driven Approach to Humor Distillation Ravi, Sahithya Huber, Patrick Shrivastava, Akshat Sagar, Aditya Aly, Ahmed Shwartz, Vered Einolghozati, Arash Computation and Language Artificial Intelligence The emergence of Large Language Models (LLMs) has brought to light promising language generation capabilities, particularly in performing tasks like complex reasoning and creative writing. Consequently, distillation through imitation of teacher responses has emerged as a popular technique to transfer knowledge from LLMs to more accessible, Small Language Models (SLMs). While this works well for simpler tasks, there is a substantial performance gap on tasks requiring intricate language comprehension and creativity, such as humor generation. We hypothesize that this gap may stem from the fact that creative tasks might be hard to learn by imitation alone and explore whether an approach, involving supplementary guidance from the teacher, could yield higher performance. To address this, we study the effect of assigning a dual role to the LLM - as a "teacher" generating data, as well as a "critic" evaluating the student's performance. Our experiments on humor generation reveal that the incorporation of feedback significantly narrows the performance gap between SLMs and their larger counterparts compared to merely relying on imitation. As a result, our research highlights the potential of using feedback as an additional dimension to data when transferring complex language abilities via distillation. |
| title | Small But Funny: A Feedback-Driven Approach to Humor Distillation |
| topic | Computation and Language Artificial Intelligence |
| url | https://arxiv.org/abs/2402.18113 |