Datasets:

nrizwan
/

FBHM

@@ -68,7 +68,7 @@ Accepted at **EMNLP 2026 Main** 🎉
 </p>
 <p align="center">
-Left: suite of 5,000 FBHM memes spread across 25 functionalities.<br>Each tile presents the functionality number, its description and the corresponding number of memes in that functionality.<br>Right: examples of constructing ten memes for ten target communities using one base image.
 </p>
 <div align="center">
@@ -78,6 +78,14 @@ Left: suite of 5,000 FBHM memes spread across 25 functionalities.<br>Each tile p
 </div>
 ------------------------------------------
 ## Abstract
 Hateful meme detection remains a formidable challenge for vision-language models, as existing benchmarks are structurally observational-confounding rhetorical hate mechanisms with target community features and preventing causal evaluation of model vulnerabilities. To address this, we introduce FBHM, a systematically curated benchmark of **F**unctionality **B**ased **H**ateful **M**emes constructed along two orthogonal axes: 25 distinct rhetorical functionalities and 10 target communities (5,000 memes total). Benchmarking state-of-the-art VLMs reveals a severe generalization gap: models highly accurate on standard datasets catastrophically drop to near-random performance on FBHM, proving they exploit dataset-specific heuristics rather than robust multimodal reasoning. To efficiently close this gap, we propose LSV (**l**earnable **s**teering **v**ectors), an ultra-low data regime strategy that applies a causal intervention objective on as few as 500 steering samples (50 unique base memes), boosting FBHM performance by ~30 Macro-F1 points while outperforming in-context learning and PEFT without degrading source-domain performance.
@@ -87,13 +95,13 @@ Hateful meme detection remains a formidable challenge for vision-language models
 FBHM can be loaded directly using the Hugging Face `datasets` library.
-### Installation
 ```bash
 pip install datasets pillow
 ```
-### Load the Dataset
 ```python
 from datasets import load_dataset
@@ -108,7 +116,7 @@ test_data = dataset["test"]
 print(dataset)
 ```
-### Inspect a Sample
 ```python
 # Select a sample from the training split
@@ -120,9 +128,7 @@ print("Label:", sample["label"])
 print("Functionality:", sample["functionality"])
 ```
-### Dataset Fields
-Each sample contains the following fields:
 | Field           | Description                                     |
 | --------------- | ----------------------------------------------- |

 </p>
 <p align="center">
+Left: suite of 5,000 FBHM memes spread across 25 functionalities. Each tile presents the functionality number, its description and the corresponding number of memes in that functionality. Right: examples of constructing ten memes for ten target communities using one base image.
 </p>
 <div align="center">
 </div>
+------------------------------------------
+```markdown
+**Content Warning** ⚠️
+This dataset contains hateful, offensive, and potentially disturbing multimodal content, including derogatory language and harmful stereotypes targeting protected groups.
+The content is provided solely for research purposes. Please use the dataset responsibly and with appropriate care when displaying or sharing examples.
+```
 ------------------------------------------
 ## Abstract
 Hateful meme detection remains a formidable challenge for vision-language models, as existing benchmarks are structurally observational-confounding rhetorical hate mechanisms with target community features and preventing causal evaluation of model vulnerabilities. To address this, we introduce FBHM, a systematically curated benchmark of **F**unctionality **B**ased **H**ateful **M**emes constructed along two orthogonal axes: 25 distinct rhetorical functionalities and 10 target communities (5,000 memes total). Benchmarking state-of-the-art VLMs reveals a severe generalization gap: models highly accurate on standard datasets catastrophically drop to near-random performance on FBHM, proving they exploit dataset-specific heuristics rather than robust multimodal reasoning. To efficiently close this gap, we propose LSV (**l**earnable **s**teering **v**ectors), an ultra-low data regime strategy that applies a causal intervention objective on as few as 500 steering samples (50 unique base memes), boosting FBHM performance by ~30 Macro-F1 points while outperforming in-context learning and PEFT without degrading source-domain performance.
 FBHM can be loaded directly using the Hugging Face `datasets` library.
+1. Installation
 ```bash
 pip install datasets pillow
 ```
+2. Load the Dataset
 ```python
 from datasets import load_dataset
 print(dataset)
 ```
+3. Inspect a Sample
 ```python
 # Select a sample from the training split
 print("Functionality:", sample["functionality"])
 ```
+4. Dataset Fields
 | Field           | Description                                     |
 | --------------- | ----------------------------------------------- |