LLM4HEP / plots /five_step_summary_stats.csv
ho22joshua's picture
initial commit
cfcbbc8
pair,step,success_count,agent_work_mean,agent_work_std,API_calls_mean,API_calls_std,total_price_mean,total_price_std,duration_mean,duration_std,input_tokens_mean,output_tokens_mean
GPT-5 Codex,1.0,10,29.71,20.43,3.8,1.03,0.18,0.15,138.28,82.4,3084.6,8740.4
GPT-5 Codex,2.0,7,15.51,5.66,4.43,1.51,0.37,0.14,239.52,90.58,10528.29,17226.71
GPT-5 Codex,3.0,9,17.17,7.11,5.44,1.67,0.53,0.24,337.15,166.49,15786.22,24568.22
GPT-5 Codex,4.0,9,6.2,1.69,3.44,0.88,0.04,0.01,223.69,101.66,3291.0,1510.33
GPT-5 Codex,5.0,5,20.85,12.88,5.8,1.1,0.35,0.26,306.1,152.33,8801.2,16389.0
GPT-5 Mini (2025-08-07),1.0,11,103.92,11.06,7.0,0.0,0.1,0.01,330.62,76.77,16302.55,23915.36
GPT-5 Mini (2025-08-07),2.0,5,30.58,4.01,7.0,0.0,0.12,0.02,354.82,68.27,25173.4,27960.2
GPT-5 Mini (2025-08-07),3.0,9,20.2,0.69,7.0,0.0,0.1,0.0,316.01,38.12,23718.78,23161.67
GPT-5 Mini (2025-08-07),4.0,10,28.16,3.11,7.0,0.0,0.04,0.01,471.36,32.86,10257.8,8990.8
GPT-5 Mini (2025-08-07),5.0,10,25.21,1.89,7.0,0.0,0.07,0.01,338.15,18.41,14457.3,15501.9
GPT-OSS-120B,1.0,54,12.57,6.99,3.37,1.03,0.0,0.0,21.24,11.84,3410.41,2853.93
GPT-OSS-120B,2.0,15,12.23,6.96,4.6,2.03,0.0,0.0,66.27,44.68,13648.13,9816.93
GPT-OSS-120B,3.0,51,9.31,3.96,4.61,1.5,0.0,0.0,72.93,29.95,14005.53,9925.16
GPT-OSS-120B,4.0,63,11.27,6.03,4.75,1.74,0.0,0.0,209.88,103.57,6150.71,3126.57
GPT-OSS-120B,5.0,60,9.57,4.3,4.73,1.62,0.0,0.0,93.18,40.95,8075.18,5187.27
Gemini 2.5 Flash,1.0,8,21.53,7.32,3.5,0.93,0.04,0.01,44.67,12.88,4281.12,6576.5
Gemini 2.5 Flash,2.0,3,19.11,10.41,5.0,2.0,0.12,0.07,134.19,67.57,17629.33,22355.67
Gemini 2.5 Flash,3.0,9,10.86,3.59,3.22,0.67,0.1,0.04,110.95,32.06,14256.33,18029.89
Gemini 2.5 Flash,4.0,9,7.67,1.72,3.0,0.0,0.02,0.01,174.12,12.04,4807.56,2845.56
Gemini 2.5 Flash,5.0,5,27.67,23.6,5.4,1.67,0.16,0.16,239.42,205.47,12244.0,30325.6
Gemini 2.5 Pro,1.0,10,21.54,1.08,3.0,0.0,0.15,0.01,85.62,13.56,3332.2,7272.6
Gemini 2.5 Pro,2.0,5,15.45,9.32,4.2,1.79,0.42,0.28,203.8,114.19,12820.0,19547.2
Gemini 2.5 Pro,3.0,10,12.51,5.54,4.0,1.7,0.45,0.19,216.06,68.87,15538.9,20521.3
Gemini 2.5 Pro,4.0,10,10.91,0.96,3.0,0.0,0.12,0.01,247.46,100.84,4531.6,5594.2
Gemini 2.5 Pro,5.0,7,11.71,5.27,3.29,0.76,0.24,0.11,245.13,140.77,7157.29,11230.86
Grok-3,1.0,10,20.47,6.48,4.8,1.14,0.13,0.04,89.64,28.87,4422.6,3522.3
Grok-3,2.0,6,9.5,4.24,4.33,1.63,0.24,0.11,164.77,82.18,11228.67,5916.67
Grok-3,3.0,9,11.61,3.0,6.11,1.45,0.41,0.11,272.38,81.44,16853.33,10242.67
Grok-3,4.0,10,8.48,4.36,4.0,1.41,0.08,0.04,179.21,76.95,4100.6,1892.0
Grok-3,5.0,1,16.91,,7.0,,0.3,,261.68,,11842.0,7772.0
O3 (2025-04-16),1.0,19,13.96,7.67,3.53,1.12,0.06,0.03,53.0,26.99,2565.05,2905.89
O3 (2025-04-16),2.0,12,7.99,4.63,3.67,1.56,0.14,0.08,113.5,66.09,8249.42,6453.42
O3 (2025-04-16),3.0,4,12.03,4.57,6.0,2.0,0.27,0.1,218.26,58.79,15121.75,12987.0
O3 (2025-04-16),4.0,20,6.3,2.17,3.2,0.62,0.04,0.01,223.55,104.37,3035.1,1501.6
O3 (2025-04-16),5.0,13,9.15,5.61,4.38,1.89,0.11,0.06,222.75,170.42,5893.69,5141.54
O4 Mini (2025-04-16),1.0,9,21.44,11.42,4.33,1.73,0.05,0.03,64.15,80.02,3085.11,5285.78
O4 Mini (2025-04-16),2.0,6,11.41,3.19,4.33,1.03,0.11,0.03,81.17,14.35,10118.5,10389.0
O4 Mini (2025-04-16),3.0,8,8.8,5.02,4.5,2.07,0.11,0.06,200.31,318.54,11173.75,10194.62
O4 Mini (2025-04-16),4.0,10,7.79,2.35,3.2,0.63,0.03,0.01,224.63,266.14,3020.9,2597.3
O4 Mini (2025-04-16),5.0,1,5.83,,3.0,,0.04,,65.78,,3746.0,3859.0
Qwen-3 (235B),1.0,10,12.96,6.93,4.0,1.41,0.01,0.01,31.98,21.7,3646.9,2457.9
Qwen-3 (235B),2.0,7,12.07,4.69,5.57,1.51,0.05,0.02,103.66,33.28,15497.29,8631.43
Qwen-3 (235B),3.0,8,14.2,1.09,7.0,0.0,0.08,0.0,167.13,41.33,24811.12,13434.75
Qwen-3 (235B),4.0,10,5.36,1.85,3.4,0.84,0.01,0.0,225.46,130.23,3784.0,1271.9
Qwen-3 (235B),5.0,2,12.19,1.15,7.0,0.0,0.04,0.0,375.09,42.65,10558.0,6912.5