Dataset · LCZZZZ
WorldMemArena long-term memory multimodal benchmark
LCZZZZ/WorldMemArena
English image and QA benchmark for testing AI agent memory over long sessions.
What's inside
Contains multi-session interaction scenarios featuring images, key memory points, and English question-answer pairs. Data is split into domains including GUI agents, embodied agents, and long-term projects, with a total volume between 10,000 and 100,000 records. File format is not specified in the description.
What you can do with it
- Evaluating long-term memory mechanisms in multimodal LLMs and AI agents
- Testing GUI and embodied agents on context retention and updating
- Tasks
- Question answering, Visual question answering
- Languages
- English
- Size
- 10K–100K rows
- Downloads this month
- 24.1K
- Likes
- 3
How to load it
from datasets import load_dataset
ds = load_dataset("LCZZZZ/WorldMemArena")