Our new X account is live! Follow @wizwand_team for updates
Home
/
Benchmarks
Audio-Visual Scene-aware Dialog on AVSD (val)
Loading...
59.48
ASR (%)
Combined Loss
40.812
45.6585
50.505
55.3515
Jan 20, 2026
ASR (%)
Updated 4d ago
Evaluation Results
Method
Method
Links
ASR (%)
Combined Loss
Attack Objective=L(com...
2026.01
59.48
Encoder-Based Cosine Similarity Loss
Attack Objective=L(cos...
2026.01
57.03
Negative Language Modeling Loss
Attack Objective=L_neg...
2026.01
50.43
Vision Attention Suppression Loss
Attack Objective=L(vis...
2026.01
49.34
Attention Randomization Loss
Attack Objective=L(ran...
2026.01
45.67
Audio Attention Amplification Loss
Attack Objective=L(aud...
2026.01
43.85
Hidden-State Similarity Loss
Attack Objective=L(hid...
2026.01
41.53
Feedback
Search any
task
Search any
task