List your company

Navigation Question Answering on RAVEN-QA Real-World Robot Tasks

97.14Overall AccuracyRAVEN (SigLIP)
1.064826.007450.9575.8926Jun 23, 2026

Evaluation results

MethodLinks
RAVEN (SigLIP)VLM Backbone=GPT-5.22026.0697.14
RAVEN (QQMM-v2)VLM Backbone=GPT-5.22026.0693.33
RAVEN (Seed)VLM Backbone=GPT-5.22026.0693.33
VLM OnlyVLM Backbone=GPT-5.22026.0692.86
RAVEN (QQMM-v2)VLM Backbone=Gemini-2.5-Flash2026.0692.38
RAVEN (Seed)VLM Backbone=Gemini-3-Pro2026.0690.47
RAVEN (QQMM-v2)VLM Backbone=Gemini-3-Pro2026.0690
RAVEN (SigLIP)VLM Backbone=Gemini-3-Pro2026.0690
VLM OnlyVLM Backbone=Gemini-3-Pro2026.0686.19
RAVEN (Seed)VLM Backbone=Gemini-2.5-Flash2026.0685.71
VLM OnlyVLM Backbone=Gemini-2.5-Flash2026.0685.24
RAVEN (SigLIP)VLM Backbone=Gemini-2.5-Flash2026.0684.76
ReMEmbR (QQMM-v2)VLM Backbone=Gemini-3-Pro2026.0680.48
ReMEmbR (QQMM-v2)VLM Backbone=Gemini-2.5-Flash2026.0676.67
ReMEmbR (QQMM-v2)VLM Backbone=GPT-5.22026.0674.76
RAVEN (QQMM-v2)VLM Backbone=Gemma3-27b2026.0667.62
RAVEN (QQMM-v2)Embedder Only=true2026.0666.67
RAVEN (Seed)Embedder Only=true2026.0666.67
RAVEN (SigLIP)VLM Backbone=Gemma3-27b2026.0647.62
RAVEN (Seed)VLM Backbone=Gemma3-27b2026.0645.71
RAVEN (SigLIP)Embedder Only=true2026.0642.86
RAVEN (SigLIP)VLM Backbone=Qwen3-VL-32b2026.0635.24
RAVEN (Seed)VLM Backbone=Qwen3-VL-32b2026.0631.9
RAVEN (QQMM-v2)VLM Backbone=Qwen3-VL-32b2026.0630
ReMEmbR (QQMM-v2)VLM Backbone=Qwen3-VL-32b2026.0628.57
ReMEmbR (QQMM-v2)VLM Backbone=Gemma3-27b2026.0614.76
VLM OnlyVLM Backbone=Gemma3-27b2026.0614.29
VLM OnlyVLM Backbone=Qwen3-VL-32b2026.064.76