Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning
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_ | 1866916529346445312 |
|---|---|
| author | Zhang, Jifan Jain, Lalit Guo, Yang Chen, Jiayi Zhou, Kuan Lok Suresh, Siddharth Wagenmaker, Andrew Sievert, Scott Rogers, Timothy Jamieson, Kevin Mankoff, Robert Nowak, Robert |
| author_facet | Zhang, Jifan Jain, Lalit Guo, Yang Chen, Jiayi Zhou, Kuan Lok Suresh, Siddharth Wagenmaker, Andrew Sievert, Scott Rogers, Timothy Jamieson, Kevin Mankoff, Robert Nowak, Robert |
| contents | We present a novel multimodal preference dataset for creative tasks, consisting of over 250 million human ratings on more than 2.2 million captions, collected through crowdsourcing rating data for The New Yorker's weekly cartoon caption contest over the past eight years. This unique dataset supports the development and evaluation of multimodal large language models and preference-based fine-tuning algorithms for humorous caption generation. We propose novel benchmarks for judging the quality of model-generated captions, utilizing both GPT4 and human judgments to establish ranking-based evaluation strategies. Our experimental results highlight the limitations of current fine-tuning methods, such as RLHF and DPO, when applied to creative tasks. Furthermore, we demonstrate that even state-of-the-art models like GPT4 and Claude currently underperform top human contestants in generating humorous captions. As we conclude this extensive data collection effort, we release the entire preference dataset to the research community, fostering further advancements in AI humor generation and evaluation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2406_10522 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning Zhang, Jifan Jain, Lalit Guo, Yang Chen, Jiayi Zhou, Kuan Lok Suresh, Siddharth Wagenmaker, Andrew Sievert, Scott Rogers, Timothy Jamieson, Kevin Mankoff, Robert Nowak, Robert Machine Learning Artificial Intelligence Computation and Language We present a novel multimodal preference dataset for creative tasks, consisting of over 250 million human ratings on more than 2.2 million captions, collected through crowdsourcing rating data for The New Yorker's weekly cartoon caption contest over the past eight years. This unique dataset supports the development and evaluation of multimodal large language models and preference-based fine-tuning algorithms for humorous caption generation. We propose novel benchmarks for judging the quality of model-generated captions, utilizing both GPT4 and human judgments to establish ranking-based evaluation strategies. Our experimental results highlight the limitations of current fine-tuning methods, such as RLHF and DPO, when applied to creative tasks. Furthermore, we demonstrate that even state-of-the-art models like GPT4 and Claude currently underperform top human contestants in generating humorous captions. As we conclude this extensive data collection effort, we release the entire preference dataset to the research community, fostering further advancements in AI humor generation and evaluation. |
| title | Humor in AI: Massive Scale Crowd-Sourced Preferences and Benchmarks for Cartoon Captioning |
| topic | Machine Learning Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2406.10522 |