Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.13608 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911709594124288 |
|---|---|
| author | Selvanayagam, Niruthiha Kurti, Ted |
| author_facet | Selvanayagam, Niruthiha Kurti, Ted |
| contents | As Large Multimodal Models (LMMs) become integral to daily digital life, understanding their safety architectures is a critical problem for AI Alignment. This paper presents a systematic analysis of OpenAI's GPT-4o mini, a globally deployed model, on the difficult task of multimodal hate speech detection. Using the Hateful Memes Challenge dataset, we conduct a multi-phase investigation on 500 samples to probe the model's reasoning and failure modes. Our central finding is the experimental identification of a "Unimodal Bottleneck," an architectural flaw where the model's advanced multimodal reasoning is systematically preempted by context-blind safety filters. A quantitative validation of 144 content policy refusals reveals that these overrides are triggered in equal measure by unimodal visual 50% and textual 50% content. We further demonstrate that this safety system is brittle, blocking not only high-risk imagery but also benign, common meme formats, leading to predictable false positives. These findings expose a fundamental tension between capability and safety in state-of-the-art LMMs, highlighting the need for more integrated, context-aware alignment strategies to ensure AI systems can be deployed both safely and effectively. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_13608 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection Selvanayagam, Niruthiha Kurti, Ted Machine Learning As Large Multimodal Models (LMMs) become integral to daily digital life, understanding their safety architectures is a critical problem for AI Alignment. This paper presents a systematic analysis of OpenAI's GPT-4o mini, a globally deployed model, on the difficult task of multimodal hate speech detection. Using the Hateful Memes Challenge dataset, we conduct a multi-phase investigation on 500 samples to probe the model's reasoning and failure modes. Our central finding is the experimental identification of a "Unimodal Bottleneck," an architectural flaw where the model's advanced multimodal reasoning is systematically preempted by context-blind safety filters. A quantitative validation of 144 content policy refusals reveals that these overrides are triggered in equal measure by unimodal visual 50% and textual 50% content. We further demonstrate that this safety system is brittle, blocking not only high-risk imagery but also benign, common meme formats, leading to predictable false positives. These findings expose a fundamental tension between capability and safety in state-of-the-art LMMs, highlighting the need for more integrated, context-aware alignment strategies to ensure AI systems can be deployed both safely and effectively. |
| title | Is GPT-4o mini Blinded by its Own Safety Filters? Exposing the Multimodal-to-Unimodal Bottleneck in Hate Speech Detection |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2509.13608 |