TwinICL Benchmark

What is TwinICL?

A diagnostic benchmark pairing text and image versions of shape tasks to compare in-context learning across modalities.

Released

2026-09-14

Evaluates

General AI, Multimodal

Openness

unknown

Importer

Claire Radar

Review status

ai-name-audit-deferred

Reported scores

0

Source provenance

• Original evidence https://arxiv.org/abs/2609.15028

TwinICL paper and code

• Paper https://arxiv.org/abs/2609.15028
• Code https://github.com/lab-flair/TwinICL

Which sources cite TwinICL?

• TwinICL

Related General AI, Multimodal benchmarks

• $\tau^\tau$-Bench: An Environment for End-To-End, Realistic Agent Construction
• 2d-fet-bench
• 3DCodeBench
• 5-Dialects-BN
• A Real-World CAD Segmentation Benchmark
• A-Bench
• A-OKVQA
• A3S-Bench