List your company

trajectory-based action prediction on AgentNetBench

31.45Per-step Action ScoreFull Cache
-0.38447.880316.14524.4097Jun 1, 2026

Evaluation results

MethodLinks
Full CacheModel=OpenCUA-7B, KV-cache budget=Full Cache2026.0631.4551.8
STaR-KVModel=OpenCUA-7B, KV-cache budget=40%2026.0625.6450.04
PyramidKVModel=OpenCUA-7B, KV-cache budget=40%2026.0625.1849.47
SnapKVModel=OpenCUA-7B, KV-cache budget=40%2026.0625.1849.32
STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.0621.149.94
GUIKVModel=OpenCUA-7B, KV-cache budget=40%2026.0620.6747.31
Full CacheModel=UI-TARS-1.5-7B, KV-cache budget=Full Cache2026.0620.3449.75
GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.0620.0248.92
STaR-KVModel=OpenCUA-7B, KV-cache budget=20%2026.0618.147.01
PyramidKVModel=OpenCUA-7B, KV-cache budget=20%2026.0618.0445.3
SnapKVModel=OpenCUA-7B, KV-cache budget=20%2026.0618.0441.17
SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.0617.4447.22
STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.0616.4547.31
GUIKVModel=OpenCUA-7B, KV-cache budget=20%2026.0615.2343.03
GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.0615.0646.59
SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.0611.0445.09
PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=40%2026.0610.345.45
STaR-KVModel=OpenCUA-7B, KV-cache budget=10%2026.0610.2642.67
GUIKVModel=OpenCUA-7B, KV-cache budget=10%2026.069.3939.16
PyramidKVModel=OpenCUA-7B, KV-cache budget=10%2026.069.1439.65
SnapKVModel=OpenCUA-7B, KV-cache budget=10%2026.067.2938.27
GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.06540.7
STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.064.640.72
SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.064.1140.04
PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=20%2026.063.5239.46
STaR-KVModel=OpenCUA-7B, KV-cache budget=5%2026.063.4235.89
GUIKVModel=OpenCUA-7B, KV-cache budget=5%2026.062.9233.78
PyramidKVModel=OpenCUA-7B, KV-cache budget=5%2026.062.8132.72
SnapKVModel=OpenCUA-7B, KV-cache budget=5%2026.062.8133.62
PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=10%2026.061.8433.12
STaR-KVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.061.332.74
GUIKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.061.2532.63
SnapKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.061.2332.04
PyramidKVModel=UI-TARS-1.5-7B, KV-cache budget=5%2026.060.8424.7