Hateful Memes Challenge
Models decide whether a meme is hateful by interpreting its image and caption together. Paired examples change one modality while preserving the other.
- Made by
- Douwe Kiela and colleagues
- Released
- May 2020
- Kind
- Bias, fairness and politics
- Level
- Everyday
- Scored
- Checked automatically
- Answers
- Multiple choice
- Tasks
- All public
- Measures
- Capability
- Risk
- Hate
- Language
- English
- Interaction
- Single turn
- Refresh
- One fixed set
- Run by
- Anyone
- In views
- Vision, Content safety
extended VL-BERT + UNITER-ITM + VILLA-ITM + ERNIE-Vil tops the leaderboard with 84.5%
Best score per entry from DrivenData’s leaderboard, read on November 20th, 2026. Where a model ran with several settings or harnesses, we keep its best. Names in gray aren't in our model list: fine-tunes, agent systems, products and older models. They don't count toward leaders.
- 1 extended VL-BERT + UNITER-ITM + VILLA-ITM + ERNIE-Vil Averaged predictions with rule-based postprocessing 84.5%
- 2 VILIO VILIO ensemble 83.1%
- 3 VisualBERT Majority voting 81.1%
- 4 UNITER-base Weighted cross-validation ensemble 80.5%
- 5 UNITER 12-model probability average 79.4%
Source: DrivenData’s leaderboard
| # | Entry | Score | Setting | Submitted by | Date | Link | Source |
|---|---|---|---|---|---|---|---|
| 1 | extended VL-BERT + UNITER-ITM + VILLA-ITM + ERNIE-Vil | 84.5% | Averaged predictions with rule-based postprocessing | alfred lab | github.com | Official leaderboard | |
| 2 | VILIO | 83.1% | VILIO ensemble | Muennighoff | github.com | Official leaderboard | |
| 3 | VisualBERT | 81.1% | Majority voting | HateDetectron | github.com | Official leaderboard | |
| 4 | UNITER-base | 80.5% | Weighted cross-validation ensemble | kingsterdam | github.com | Official leaderboard | |
| 5 | UNITER | 79.4% | 12-model probability average | burebista | github.com | Official leaderboard |
No lab model card reports it yet
We read the headline results of 540 launch posts from 57 labs, and none of them list this benchmark.
More in Safety and alignment
- FORTRESS In use Paired adversarial and benign national-security misuse prompts. Rubric-graded risk score on adversarial and over-refusal score on benign prompts. Claude Opus 5.5 5.78%
- StrongREJECT In use Forbidden-prompt jailbreak benchmark with a rubric grader that scores how useful a non-refusal actually is.
Inkling 98.6%
- Gray Swan IPI Arena benchmark In use Indirect prompt-injection attacks from public red-teaming competitions replayed on tool, coding and browser agents. Attack success rate at k attempts.
Gemini 4 Carbon 0%
- Covert misbehavior (Thinking Machines) In use Agentic tasks that check whether a model misbehaves and conceals it. Scored by the rate of covert misbehavior.
Inkling 2 Mini 8.30%
- Destructive actions (Thinking Machines) In use Agentic tasks that check whether a model takes destructive actions it was not authorized to take. Scored by the rate of destructive actions.
Inkling 2 Small 4.70%
- Misbehavior (Thinking Machines) In use Agentic tasks that check whether a model fabricates results, ignores instructions or hides what it did when a task is hard. Scored by the rate of misbehavior.
Inkling 2 Mini 9.80%
Data audit
All 8 checks pass. Every benchmark
Catalog record 2 pass
- Pass Task access All tasks are public.
- Pass Confidence Checked against a primary source.
Scores 2 pass 3 notes
- Pass Independent runs Scores from the official leaderboard (5 entries).
- Pass Models scored 5 entries on its leaderboard, none a model we track.
- Note Newest score No scores yet.
- Note Frontier models No models scored yet.
- Note Headroom No scores yet.
Leaderboard import 3 pass
- Pass Importer We read the board on DrivenData: 5 entries at the last read, none a model we track.
- Pass Last read November 20th, 2026.
- Pass Read worked The last read worked.
Consistency 1 pass 1 note
- Note Lab claims No model has both a lab number and an independent run.
- Pass Known issues No published audit found problems.