List your company

Scene Decoding on CLEVR (test)

0Parse FailsNaive Baseline
-0.120.691.52.31May 26, 2026

Evaluation results

MethodLinks
Naive BaselineCB=N/A2026.0502.2756.83.842.22
Qwen3.5-397B-A17BCB=322026.0501.43513.221.9
Qwen3.5-397B-A17BCB=642026.050145.83.372
Qwen3.5-397B-A17BCB=1282026.0501.5350.53.171.91
DeepSeek-V3.1-vLLM-2CB=642026.0500.9354.33.641.97
DeepSeek-V3.1-vLLM-2CB=1282026.0501.1348.73.151.84
GPT-OSS-120BCB=322026.0512.3646.92.451.71
GPT-OSS-120BCB=1282026.0512.3652.44.612.13
DeepSeek-V3.1-vLLM-2CB=322026.0511.3651.33.682.11
GPT-OSS-120BCB=642026.0532.8353.63.261.87