MLE-Bench: Multimodal Model Perception Benchmark

The Multi-Level Existence Benchmark (MLE-Bench) is a dataset designed for fine-grained evaluation of multimodal models' perceptual abilities. It assesses 'pure' perception using 4-choice questions about object or scene existence within images. The dataset was created by JunlinHan and was last updated on March 8, 2026.

Use Cases

• Benchmarking model perceptual accuracy based on existence questions.
• Evaluating AI performance independent of complex reasoning as described.
• Analyzing model performance across different categories of target objects.

Strengths

• Specifically designed for fine-grained evaluation of perceptual abilities.
• Focuses on 'pure' perception independent of complex reasoning.
• Categorizes questions based on target object characteristics.

Limitations

• Description metadata is limited; actual data quality requires manual inspection after download.
• Column-level documentation is absent; field semantics must be inferred after download.
• Row count is unknown, which may limit suitability assessment.

Provenance

Source

huggingface

Freshness

Last updated 2026-03-08 06:31:19; freshness should be verified.