List your company

Interactive Navigation on CoIN-Bench Unseen (val)

17SRProCompNav
-0.683.918.513.09May 7, 2026

Evaluation results

MethodLinks
ProCompNavJudge=Comp, Interact=true2026.05176.2
Pooled Independent MatchingJudge=Indep, Interact=true2026.0513.35.1
Context-NavJudge=Indep, Interact=false2026.0511.35.2
AIUTA*Judge=Indep, Interact=true2026.058.94
AIUTAJudge=Indep, Interact=true2026.056.72.3
3D-Mem*Judge=Comp, Interact=false2026.055.73.9
PSLJudge=–, Interact=false2026.054.61.4
Monolithic-GOATJudge=–, Interact=false2026.050.20.1
VLFMJudge=Indep, Interact=false2026.0500