52 lines
1.3 KiB
JSON
52 lines
1.3 KiB
JSON
[
|
|
{
|
|
"agent": "keyword_spam",
|
|
"avg_evidence_timing": 0,
|
|
"avg_primary_reward": 0.0,
|
|
"avg_reward": 0.07300000000000001,
|
|
"avg_surface_rate": 0.0,
|
|
"avg_trigger_rate": 0.6777777777777778,
|
|
"avg_useless_ratio": 0.48000000000000004,
|
|
"episodes": 30
|
|
},
|
|
{
|
|
"agent": "present_all",
|
|
"avg_evidence_timing": 0,
|
|
"avg_primary_reward": 0.0,
|
|
"avg_reward": 0.0,
|
|
"avg_surface_rate": 0.0,
|
|
"avg_trigger_rate": 0.0,
|
|
"avg_useless_ratio": 0.0,
|
|
"episodes": 30
|
|
},
|
|
{
|
|
"agent": "random",
|
|
"avg_evidence_timing": 0,
|
|
"avg_primary_reward": 0.0,
|
|
"avg_reward": 0.0,
|
|
"avg_surface_rate": 0.0,
|
|
"avg_trigger_rate": 0.0,
|
|
"avg_useless_ratio": 1.0,
|
|
"episodes": 30
|
|
},
|
|
{
|
|
"agent": "scripted_oracle",
|
|
"avg_evidence_timing": 1.9,
|
|
"avg_primary_reward": 0.95,
|
|
"avg_reward": 0.9023333333333334,
|
|
"avg_surface_rate": 0.95,
|
|
"avg_trigger_rate": 0.95,
|
|
"avg_useless_ratio": 0.0,
|
|
"episodes": 30
|
|
},
|
|
{
|
|
"agent": "trained_sft_grpo_run2",
|
|
"avg_evidence_timing": 0.8,
|
|
"avg_primary_reward": 0.4611111111111111,
|
|
"avg_reward": 0.3868888888888889,
|
|
"avg_surface_rate": 0.4611111111111111,
|
|
"avg_trigger_rate": 0.5888888888888889,
|
|
"avg_useless_ratio": 0.8285714285714285,
|
|
"episodes": 30
|
|
}
|
|
] |