Interactive Navigation on CoIN-Bench Seen Synonyms (val)
28.1SRProCompNav
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ProCompNavJudge=Comp, Interact=true2026.05 | 28.1 | 8.5 | |
| Pooled Independent MatchingJudge=Indep, Interact=true2026.05 | 22 | 8.1 | |
| Context-NavJudge=Indep, Interact=false2026.05 | 20.3 | 10.9 | |
| 3D-Mem*Judge=Comp, Interact=false2026.05 | 18.9 | 13.6 | |
| AIUTA*Judge=Indep, Interact=true2026.05 | 15.3 | 8.4 | |
| AIUTAJudge=Indep, Interact=true2026.05 | 14.4 | 8 | |
| Monolithic-GOATJudge=–, Interact=false2026.05 | 13.1 | 6.5 | |
| PSLJudge=–, Interact=false2026.05 | 8.9 | 2.8 | |
| VLFMJudge=Indep, Interact=false2026.05 | 0 | 0 |