LU-500 Benchmark

What is LU-500?

LU-500 evaluates concept unlearning for logos with nearly 10,000 pairs, including explicit and implicit contextual tracks, and a multi-grained protocol…

Released

2026-07-27

Evaluates

General AI, Vision & 3D, cs.CV

Openness

unknown

Importer

Claire Radar

Review status

unreviewed

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2607.24101

LU-500 paper

• Paper https://arxiv.org/abs/2607.24101

Which sources cite LU-500?

• LU-500

Related General AI, Vision & 3D, cs.CV benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 360CityArena
• 3DCodeBench
• 5-Dialects-BN
• A3S-Bench
• AARRI-Bench
• ActBench
• ActiveSciBench