Coverage (30d)
0vs1
This Week
0vs0
Evidence
1 articlesRelationships
2Timeline
Tuna-22026-05-04
Introduced Tuna-2, a native unified multimodal model using pixel embeddings
GPT-4V2026-04-04
Documented failure to generate coherent world maps, becoming a benchmark for spatial reasoning weaknesses
Ecosystem
Tuna-2
competes withGPT-4V1 src
competes withGemini1 src
competes withLLaVA1 src
GPT-4V
competes withTuna-21 src
usesVariational Autoencoders1 src