Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Audio-based Reasoning on BigBench Audio

97.6Accuracy

Fun-Realtime-Audiochat

0.67225.8365176.164Jan 26, 2026Feb 21, 2026Mar 20, 2026Apr 16, 2026May 13, 2026Jun 9, 2026Jul 6, 2026
Updated 19d ago

Evaluation Results

MethodLinks
97.6--
97.6--
96.6--
93.3--
90.7--
90--
2026.07
88.1--
83.3--
81.4--
2026.01
73.7776,00084,000
73--
71.9209338
2026.01
71.633,00046,000
2026.01
71.533,00045,000
2026.07
71.3--
2026.07
68.9--
68.6--
2026.07
64.3--
2026.07
63.6--
2026.07
60.2--
59--
58.7--
56.8--
2026.01
55.2206258
2026.01
54.959903,166
2026.01
54.5213264
54.3--
51--
2026.01
43.1195.13332.92
38.8--
33.4--
19.1--
2026.07
16--
2026.07
4.4--