Vision-Language Compositionality on ColorSwap
85.17AccuracyTTM
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| TTMBackbone=SmolVLM-256M-Instruct, Scoring Algorithm=Algorithm 12025.10 | 85.17 | 5.2 | 6.5 | 25.9 | |
| SimpleMatchBackbone=SmolVLM-256M-Instruct, Scoring Algorithm=GroupMatch2025.10 | 80 | — | — | — | |
| SmolVLMBackbone=SmolVLM-256M-Instruct, Scoring Algorithm=GroupScore2025.10 | 30 | — | — | — |