Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions
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_ | 1866916623615524864 |
|---|---|
| author | Fu, Yuhan Xie, Ruobing Liu, Jiazhen Lan, Bangxiang Sun, Xingwu Kang, Zhanhui Li, Xirong |
| author_facet | Fu, Yuhan Xie, Ruobing Liu, Jiazhen Lan, Bangxiang Sun, Xingwu Kang, Zhanhui Li, Xirong |
| contents | Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2410_11701 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions Fu, Yuhan Xie, Ruobing Liu, Jiazhen Lan, Bangxiang Sun, Xingwu Kang, Zhanhui Li, Xirong Computation and Language Artificial Intelligence Computer Vision and Pattern Recognition Multimedia Hallucinations in multimodal large language models (MLLMs) hinder their practical applications. To address this, we propose a Magnifier Prompt (MagPrompt), a simple yet effective method to tackle hallucinations in MLLMs via extremely simple instructions. MagPrompt is based on the following two key principles, which guide the design of various effective prompts, demonstrating robustness: (1) MLLMs should focus more on the image. (2) When there are conflicts between the image and the model's inner knowledge, MLLMs should prioritize the image. MagPrompt is training-free and can be applied to open-source and closed-source models, such as GPT-4o and Gemini-pro. It performs well across many datasets and its effectiveness is comparable or even better than more complex methods like VCD. Furthermore, our prompt design principles and experimental analyses provide valuable insights into multimodal hallucination. |
| title | Magnifier Prompt: Tackling Multimodal Hallucination via Extremely Simple Instructions |
| topic | Computation and Language Artificial Intelligence Computer Vision and Pattern Recognition Multimedia |
| url | https://arxiv.org/abs/2410.11701 |