EASI Leaderboard

Submit a Model

Authenticate with HuggingFace, then fill in each field below. The form expands as you go. If you are continuously facing difficulty uploading, please email your eval results to easi-lmms-lab@outlook.com.

Authenticate so that we can identify you and your evaluation results.

Model Configuration

Check benchmarks you evaluated. Unchecked = not evaluated (null). At least one required.

VSI-Bench
Sub-scores for VSI-Bench
MMSI-Bench
Sub-scores for MMSI-Bench
MindCube-Tiny
Sub-scores for MindCube-Tiny
ViewSpatial
Sub-scores for ViewSpatial
SITE
Sub-scores for SITE
BLINK
Sub-scores for BLINK
3DSRBench
Sub-scores for 3DSRBench
EmbSpatial
Sub-scores for EmbSpatial
MMSI-Video-Bench
Sub-scores for MMSI-Video-Bench
OmniSpatial (Manual CoT)
Sub-scores for OmniSpatial (Manual CoT)
SPAR-Bench
Sub-scores for SPAR-Bench
VSI-Debiased
Sub-scores for VSI-Debiased

Upload a .zip file containing all evaluation result files for verification.

.ZIP
↑

Upload Evaluation Results

Drag & drop your .zip file here, or click to browse

ZIP only · Max 20 MB · Required

Review & Submit

Submitted by:Model:Type:Precision:Revision:Weight:Backend:
Scores
VSI-Bench:null
MMSI-Bench:null
MindCube-Tiny:null
ViewSpatial:null
SITE:null
BLINK:null
3DSRBench:null
EmbSpatial:null
MMSI-Video-Bench:null
OmniSpatial (Manual CoT):null
SPAR-Bench:null
VSI-Debiased:null