MINED (ACL26 🔥)
Collection
MINED introduces a comprehensive benchmark designed to evaluate temporal awareness in Large Multimodal Models and investigate knowledge editing for up • 2 items • Updated • 2
Need help to make the dataset viewer work? Make sure to review how to configure the dataset viewer, and open a discussion for direct support.
Temporal Awareness Evaluation, Comprehensive Benchmarking, and Multi-Dimensional Analysis!
MINED
|--
inference_data (json/jsonl)
| |-- Dimension1_time_agnostic.json
| |-- Dimension1_temporal_interval.json
| |-- Dimension1_time_agnostic.json
| |-- Dimension2_awareness_future.json
| |-- Dimension2_awareness_past.json
| |-- Dimension3_future_unanswerable_date.json
| |-- Dimension3_previous_unanswerable_date.json
| |-- Dimension4_understanding.json
| |-- Dimension5_calculation.json
| |-- Dimension5_ranking.json
| |-- Dimension6_robustness.json
|-- imgs
| |-- MINED_Image.zip