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

Visual Reasoning on TIR-Bench

51.8Average Score

InterSketch-8B

15.50424.92734.3543.773May 26, 2026
Updated 7d ago

Evaluation Results

MethodLinks
2026.05
51.885527121.27597.359.228.463.34232
2026.05
4970.849432133.358.35529.965.86466
2026.05
4642.5556416.453.377.331.74157.56642
2026.05
30.750.84558.811.734.732.534.942.52840
2026.05
2924.2441210.42530.721.728.558.33442
2026.05
26.525.835113.45317.321.723.559.21232
2026.05
24.313.334146.86514.717.519.954.21638
2026.05
22.327.53144.848.3201013.755.81628
2026.05
20.822.52904.65013.36.512.954.242.916
2026.05
19.825362516.6201520.636.72424
2026.05
18.311.731113.73.317.317.511352634
2026.05
17.816.72213.941.7126.719.950.819.916
2026.05
17.317.52574.81.79.322.517.332.52226
2026.05
17.3202606.2102022.519.4351026
2026.05
16.933.32324.5017.311.716.636.7622