Stereo_Depth / analysis /intervention_analysis /analysis_summary.json
LLDDSS's picture
Upload folder using huggingface_hub
18104f9 verified
Raw History Blame Contribute Delete
150 kB
{
"result_dir": "/HDD2/disheng/Spatial_VLM_Stereo/June/Spatial_VLM_Stereo/01_3_prelimilary_intervention_different_baseline/result",
"runs": [
{
"model_id": "Qwen/Qwen2.5-VL-3B-Instruct",
"task": "vsr",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 2195,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 2195,
"accuracy": 0.7712984054669704,
"num_failures": 502,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 2195,
"accuracy": 0.7676537585421412,
"num_failures": 510,
"flip_rate_vs_reference": 0.07015945330296128,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 2195,
"accuracy": 0.7594533029612757,
"num_failures": 528,
"flip_rate_vs_reference": 0.07927107061503418,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 2195,
"accuracy": 0.7580865603644646,
"num_failures": 531,
"flip_rate_vs_reference": 0.08792710706150342,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 2195,
"accuracy": 0.752619589977221,
"num_failures": 543,
"flip_rate_vs_reference": 0.09977220956719818,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 2195,
"accuracy": 0.7480637813211846,
"num_failures": 553,
"flip_rate_vs_reference": 0.10888382687927108,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.02323462414578581,
"stability": {
"always_correct": 1434,
"always_wrong": 332,
"unstable": 429,
"always_correct_rate": 0.6533029612756264,
"always_wrong_rate": 0.15125284738041003,
"unstable_rate": 0.19544419134396354
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 502,
"failures_b": 510,
"failures_shared": 429,
"failure_jaccard": 0.7358490566037735,
"failure_recall_a_in_b": 0.8545816733067729,
"successes_a": 1693,
"successes_b": 1685,
"successes_shared": 1612,
"success_jaccard": 0.912797281993205,
"answer_agreement": 0.9298405466970388
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 502,
"failures_b": 528,
"failures_shared": 428,
"failure_jaccard": 0.7109634551495017,
"failure_recall_a_in_b": 0.852589641434263,
"successes_a": 1693,
"successes_b": 1667,
"successes_shared": 1593,
"success_jaccard": 0.9015280135823429,
"answer_agreement": 0.9207289293849659
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 502,
"failures_b": 531,
"failures_shared": 420,
"failure_jaccard": 0.6851549755301795,
"failure_recall_a_in_b": 0.8366533864541833,
"successes_a": 1693,
"successes_b": 1664,
"successes_shared": 1582,
"success_jaccard": 0.8912676056338028,
"answer_agreement": 0.9120728929384966
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 502,
"failures_b": 543,
"failures_shared": 413,
"failure_jaccard": 0.6534810126582279,
"failure_recall_a_in_b": 0.8227091633466136,
"successes_a": 1693,
"successes_b": 1652,
"successes_shared": 1563,
"success_jaccard": 0.877104377104377,
"answer_agreement": 0.9002277904328019
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 502,
"failures_b": 553,
"failures_shared": 408,
"failure_jaccard": 0.6306027820710973,
"failure_recall_a_in_b": 0.8127490039840638,
"successes_a": 1693,
"successes_b": 1642,
"successes_shared": 1548,
"success_jaccard": 0.8662562954672636,
"answer_agreement": 0.8911161731207289
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 510,
"failures_b": 528,
"failures_shared": 433,
"failure_jaccard": 0.715702479338843,
"failure_recall_a_in_b": 0.8490196078431372,
"successes_a": 1685,
"successes_b": 1667,
"successes_shared": 1590,
"success_jaccard": 0.9023836549375709,
"answer_agreement": 0.9216400911161731
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 510,
"failures_b": 531,
"failures_shared": 434,
"failure_jaccard": 0.71499176276771,
"failure_recall_a_in_b": 0.8509803921568627,
"successes_a": 1685,
"successes_b": 1664,
"successes_shared": 1588,
"success_jaccard": 0.9017603634298694,
"answer_agreement": 0.9211845102505695
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 510,
"failures_b": 543,
"failures_shared": 429,
"failure_jaccard": 0.6875,
"failure_recall_a_in_b": 0.8411764705882353,
"successes_a": 1685,
"successes_b": 1652,
"successes_shared": 1571,
"success_jaccard": 0.8895809739524349,
"answer_agreement": 0.9111617312072893
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 510,
"failures_b": 553,
"failures_shared": 423,
"failure_jaccard": 0.6609375,
"failure_recall_a_in_b": 0.8294117647058824,
"successes_a": 1685,
"successes_b": 1642,
"successes_shared": 1555,
"success_jaccard": 0.8775395033860045,
"answer_agreement": 0.9011389521640091
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 528,
"failures_b": 531,
"failures_shared": 445,
"failure_jaccard": 0.7247557003257329,
"failure_recall_a_in_b": 0.8428030303030303,
"successes_a": 1667,
"successes_b": 1664,
"successes_shared": 1581,
"success_jaccard": 0.9034285714285715,
"answer_agreement": 0.9230068337129841
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 528,
"failures_b": 543,
"failures_shared": 441,
"failure_jaccard": 0.7,
"failure_recall_a_in_b": 0.8352272727272727,
"successes_a": 1667,
"successes_b": 1652,
"successes_shared": 1565,
"success_jaccard": 0.8922462941847207,
"answer_agreement": 0.9138952164009112
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 528,
"failures_b": 553,
"failures_shared": 430,
"failure_jaccard": 0.6605222734254992,
"failure_recall_a_in_b": 0.8143939393939394,
"successes_a": 1667,
"successes_b": 1642,
"successes_shared": 1544,
"success_jaccard": 0.8747875354107649,
"answer_agreement": 0.8993166287015946
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 531,
"failures_b": 543,
"failures_shared": 447,
"failure_jaccard": 0.7129186602870813,
"failure_recall_a_in_b": 0.8418079096045198,
"successes_a": 1664,
"successes_b": 1652,
"successes_shared": 1568,
"success_jaccard": 0.897025171624714,
"answer_agreement": 0.9179954441913439
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 531,
"failures_b": 553,
"failures_shared": 445,
"failure_jaccard": 0.6964006259780907,
"failure_recall_a_in_b": 0.8380414312617702,
"successes_a": 1664,
"successes_b": 1642,
"successes_shared": 1556,
"success_jaccard": 0.8891428571428571,
"answer_agreement": 0.911617312072893
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 543,
"failures_b": 553,
"failures_shared": 460,
"failure_jaccard": 0.7232704402515723,
"failure_recall_a_in_b": 0.8471454880294659,
"successes_a": 1652,
"successes_b": 1642,
"successes_shared": 1559,
"success_jaccard": 0.8985590778097983,
"answer_agreement": 0.9198177676537586
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-3B-Instruct",
"task": "youtube_level1",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 821,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 821,
"accuracy": 0.3995127892813642,
"num_failures": 493,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 821,
"accuracy": 0.4056029232643118,
"num_failures": 488,
"flip_rate_vs_reference": 0.14859926918392205,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 821,
"accuracy": 0.41169305724725946,
"num_failures": 483,
"flip_rate_vs_reference": 0.1559074299634592,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 821,
"accuracy": 0.4104750304506699,
"num_failures": 484,
"flip_rate_vs_reference": 0.16321559074299635,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 821,
"accuracy": 0.3995127892813642,
"num_failures": 493,
"flip_rate_vs_reference": 0.1814859926918392,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 821,
"accuracy": 0.4007308160779537,
"num_failures": 492,
"flip_rate_vs_reference": 0.19366626065773446,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.012180267965895275,
"stability": {
"always_correct": 247,
"always_wrong": 391,
"unstable": 183,
"always_correct_rate": 0.30085261875761266,
"always_wrong_rate": 0.4762484774665043,
"unstable_rate": 0.22289890377588306
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 493,
"failures_b": 488,
"failures_shared": 450,
"failure_jaccard": 0.847457627118644,
"failure_recall_a_in_b": 0.9127789046653144,
"successes_a": 328,
"successes_b": 333,
"successes_shared": 290,
"success_jaccard": 0.7816711590296496,
"answer_agreement": 0.8514007308160779
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 493,
"failures_b": 483,
"failures_shared": 451,
"failure_jaccard": 0.8590476190476191,
"failure_recall_a_in_b": 0.9148073022312373,
"successes_a": 328,
"successes_b": 338,
"successes_shared": 296,
"success_jaccard": 0.8,
"answer_agreement": 0.8440925700365408
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 493,
"failures_b": 484,
"failures_shared": 450,
"failure_jaccard": 0.8538899430740038,
"failure_recall_a_in_b": 0.9127789046653144,
"successes_a": 328,
"successes_b": 337,
"successes_shared": 294,
"success_jaccard": 0.7924528301886793,
"answer_agreement": 0.8367844092570037
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 493,
"failures_b": 493,
"failures_shared": 445,
"failure_jaccard": 0.822550831792976,
"failure_recall_a_in_b": 0.9026369168356998,
"successes_a": 328,
"successes_b": 328,
"successes_shared": 280,
"success_jaccard": 0.7446808510638298,
"answer_agreement": 0.8185140073081608
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 493,
"failures_b": 492,
"failures_shared": 441,
"failure_jaccard": 0.8106617647058824,
"failure_recall_a_in_b": 0.8945233265720081,
"successes_a": 328,
"successes_b": 329,
"successes_shared": 277,
"success_jaccard": 0.7289473684210527,
"answer_agreement": 0.8063337393422655
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 488,
"failures_b": 483,
"failures_shared": 449,
"failure_jaccard": 0.8601532567049809,
"failure_recall_a_in_b": 0.9200819672131147,
"successes_a": 333,
"successes_b": 338,
"successes_shared": 299,
"success_jaccard": 0.803763440860215,
"answer_agreement": 0.8440925700365408
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 488,
"failures_b": 484,
"failures_shared": 447,
"failure_jaccard": 0.8514285714285714,
"failure_recall_a_in_b": 0.9159836065573771,
"successes_a": 333,
"successes_b": 337,
"successes_shared": 296,
"success_jaccard": 0.7914438502673797,
"answer_agreement": 0.8392204628501827
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 488,
"failures_b": 493,
"failures_shared": 454,
"failure_jaccard": 0.8614800759013282,
"failure_recall_a_in_b": 0.930327868852459,
"successes_a": 333,
"successes_b": 328,
"successes_shared": 294,
"success_jaccard": 0.8010899182561307,
"answer_agreement": 0.833130328867235
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 488,
"failures_b": 492,
"failures_shared": 442,
"failure_jaccard": 0.8215613382899628,
"failure_recall_a_in_b": 0.9057377049180327,
"successes_a": 333,
"successes_b": 329,
"successes_shared": 283,
"success_jaccard": 0.7467018469656992,
"answer_agreement": 0.8063337393422655
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 483,
"failures_b": 484,
"failures_shared": 446,
"failure_jaccard": 0.8560460652591171,
"failure_recall_a_in_b": 0.9233954451345756,
"successes_a": 338,
"successes_b": 337,
"successes_shared": 300,
"success_jaccard": 0.8,
"answer_agreement": 0.8453105968331304
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 483,
"failures_b": 493,
"failures_shared": 450,
"failure_jaccard": 0.8555133079847909,
"failure_recall_a_in_b": 0.9316770186335404,
"successes_a": 338,
"successes_b": 328,
"successes_shared": 295,
"success_jaccard": 0.7951482479784366,
"answer_agreement": 0.8550548112058465
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 483,
"failures_b": 492,
"failures_shared": 445,
"failure_jaccard": 0.839622641509434,
"failure_recall_a_in_b": 0.9213250517598344,
"successes_a": 338,
"successes_b": 329,
"successes_shared": 291,
"success_jaccard": 0.773936170212766,
"answer_agreement": 0.8270401948842875
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 484,
"failures_b": 493,
"failures_shared": 452,
"failure_jaccard": 0.8609523809523809,
"failure_recall_a_in_b": 0.9338842975206612,
"successes_a": 337,
"successes_b": 328,
"successes_shared": 296,
"success_jaccard": 0.8021680216802168,
"answer_agreement": 0.8514007308160779
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 484,
"failures_b": 492,
"failures_shared": 441,
"failure_jaccard": 0.8242990654205608,
"failure_recall_a_in_b": 0.9111570247933884,
"successes_a": 337,
"successes_b": 329,
"successes_shared": 286,
"success_jaccard": 0.7526315789473684,
"answer_agreement": 0.8136419001218027
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 493,
"failures_b": 492,
"failures_shared": 459,
"failure_jaccard": 0.8726235741444867,
"failure_recall_a_in_b": 0.9310344827586207,
"successes_a": 328,
"successes_b": 329,
"successes_shared": 295,
"success_jaccard": 0.8149171270718232,
"answer_agreement": 0.8660170523751523
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-3B-Instruct",
"task": "youtube_level2",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 764,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 764,
"accuracy": 0.362565445026178,
"num_failures": 487,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 764,
"accuracy": 0.3743455497382199,
"num_failures": 478,
"flip_rate_vs_reference": 0.1387434554973822,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 764,
"accuracy": 0.3782722513089005,
"num_failures": 475,
"flip_rate_vs_reference": 0.17277486910994763,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 764,
"accuracy": 0.3795811518324607,
"num_failures": 474,
"flip_rate_vs_reference": 0.1793193717277487,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 764,
"accuracy": 0.38089005235602097,
"num_failures": 473,
"flip_rate_vs_reference": 0.17408376963350786,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 764,
"accuracy": 0.3704188481675393,
"num_failures": 481,
"flip_rate_vs_reference": 0.2107329842931937,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.01832460732984298,
"stability": {
"always_correct": 198,
"always_wrong": 392,
"unstable": 174,
"always_correct_rate": 0.2591623036649215,
"always_wrong_rate": 0.5130890052356021,
"unstable_rate": 0.22774869109947643
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 487,
"failures_b": 478,
"failures_shared": 450,
"failure_jaccard": 0.8737864077669902,
"failure_recall_a_in_b": 0.9240246406570842,
"successes_a": 277,
"successes_b": 286,
"successes_shared": 249,
"success_jaccard": 0.7929936305732485,
"answer_agreement": 0.8612565445026178
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 487,
"failures_b": 475,
"failures_shared": 441,
"failure_jaccard": 0.8464491362763915,
"failure_recall_a_in_b": 0.9055441478439425,
"successes_a": 277,
"successes_b": 289,
"successes_shared": 243,
"success_jaccard": 0.7523219814241486,
"answer_agreement": 0.8272251308900523
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 487,
"failures_b": 474,
"failures_shared": 437,
"failure_jaccard": 0.833969465648855,
"failure_recall_a_in_b": 0.8973305954825462,
"successes_a": 277,
"successes_b": 290,
"successes_shared": 240,
"success_jaccard": 0.7339449541284404,
"answer_agreement": 0.8206806282722513
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 487,
"failures_b": 473,
"failures_shared": 437,
"failure_jaccard": 0.8355640535372849,
"failure_recall_a_in_b": 0.8973305954825462,
"successes_a": 277,
"successes_b": 291,
"successes_shared": 241,
"success_jaccard": 0.7370030581039755,
"answer_agreement": 0.8259162303664922
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 487,
"failures_b": 481,
"failures_shared": 435,
"failure_jaccard": 0.8161350844277674,
"failure_recall_a_in_b": 0.893223819301848,
"successes_a": 277,
"successes_b": 283,
"successes_shared": 231,
"success_jaccard": 0.7021276595744681,
"answer_agreement": 0.7892670157068062
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 478,
"failures_b": 475,
"failures_shared": 444,
"failure_jaccard": 0.8722986247544204,
"failure_recall_a_in_b": 0.9288702928870293,
"successes_a": 286,
"successes_b": 289,
"successes_shared": 255,
"success_jaccard": 0.796875,
"answer_agreement": 0.8547120418848168
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 478,
"failures_b": 474,
"failures_shared": 438,
"failure_jaccard": 0.8521400778210116,
"failure_recall_a_in_b": 0.9163179916317992,
"successes_a": 286,
"successes_b": 290,
"successes_shared": 250,
"success_jaccard": 0.7668711656441718,
"answer_agreement": 0.8416230366492147
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 478,
"failures_b": 473,
"failures_shared": 441,
"failure_jaccard": 0.8647058823529412,
"failure_recall_a_in_b": 0.9225941422594143,
"successes_a": 286,
"successes_b": 291,
"successes_shared": 254,
"success_jaccard": 0.7863777089783281,
"answer_agreement": 0.8586387434554974
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 478,
"failures_b": 481,
"failures_shared": 435,
"failure_jaccard": 0.8301526717557252,
"failure_recall_a_in_b": 0.9100418410041841,
"successes_a": 286,
"successes_b": 283,
"successes_shared": 240,
"success_jaccard": 0.729483282674772,
"answer_agreement": 0.8141361256544503
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 475,
"failures_b": 474,
"failures_shared": 436,
"failure_jaccard": 0.8499025341130604,
"failure_recall_a_in_b": 0.9178947368421052,
"successes_a": 289,
"successes_b": 290,
"successes_shared": 251,
"success_jaccard": 0.7652439024390244,
"answer_agreement": 0.8390052356020943
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 475,
"failures_b": 473,
"failures_shared": 435,
"failure_jaccard": 0.847953216374269,
"failure_recall_a_in_b": 0.9157894736842105,
"successes_a": 289,
"successes_b": 291,
"successes_shared": 251,
"success_jaccard": 0.7629179331306991,
"answer_agreement": 0.8390052356020943
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 475,
"failures_b": 481,
"failures_shared": 436,
"failure_jaccard": 0.8384615384615385,
"failure_recall_a_in_b": 0.9178947368421052,
"successes_a": 289,
"successes_b": 283,
"successes_shared": 244,
"success_jaccard": 0.7439024390243902,
"answer_agreement": 0.8180628272251309
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 474,
"failures_b": 473,
"failures_shared": 436,
"failure_jaccard": 0.8532289628180039,
"failure_recall_a_in_b": 0.919831223628692,
"successes_a": 290,
"successes_b": 291,
"successes_shared": 253,
"success_jaccard": 0.7713414634146342,
"answer_agreement": 0.8534031413612565
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 474,
"failures_b": 481,
"failures_shared": 438,
"failure_jaccard": 0.8471953578336557,
"failure_recall_a_in_b": 0.9240506329113924,
"successes_a": 290,
"successes_b": 283,
"successes_shared": 247,
"success_jaccard": 0.7576687116564417,
"answer_agreement": 0.837696335078534
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 473,
"failures_b": 481,
"failures_shared": 439,
"failure_jaccard": 0.8524271844660194,
"failure_recall_a_in_b": 0.9281183932346723,
"successes_a": 291,
"successes_b": 283,
"successes_shared": 249,
"success_jaccard": 0.7661538461538462,
"answer_agreement": 0.8298429319371727
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-3B-Instruct",
"task": "youtube_level3",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 836,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 836,
"accuracy": 0.26674641148325356,
"num_failures": 613,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 836,
"accuracy": 0.2452153110047847,
"num_failures": 631,
"flip_rate_vs_reference": 0.2452153110047847,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 836,
"accuracy": 0.23923444976076555,
"num_failures": 636,
"flip_rate_vs_reference": 0.23803827751196172,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 836,
"accuracy": 0.25239234449760767,
"num_failures": 625,
"flip_rate_vs_reference": 0.26435406698564595,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 836,
"accuracy": 0.2595693779904306,
"num_failures": 619,
"flip_rate_vs_reference": 0.25478468899521534,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 836,
"accuracy": 0.26555023923444976,
"num_failures": 614,
"flip_rate_vs_reference": 0.27392344497607657,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.027511961722488015,
"stability": {
"always_correct": 111,
"always_wrong": 502,
"unstable": 223,
"always_correct_rate": 0.13277511961722488,
"always_wrong_rate": 0.6004784688995215,
"unstable_rate": 0.26674641148325356
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 613,
"failures_b": 631,
"failures_shared": 569,
"failure_jaccard": 0.8429629629629629,
"failure_recall_a_in_b": 0.9282218597063622,
"successes_a": 223,
"successes_b": 205,
"successes_shared": 161,
"success_jaccard": 0.602996254681648,
"answer_agreement": 0.7547846889952153
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 613,
"failures_b": 636,
"failures_shared": 570,
"failure_jaccard": 0.8394698085419735,
"failure_recall_a_in_b": 0.9298531810766721,
"successes_a": 223,
"successes_b": 200,
"successes_shared": 157,
"success_jaccard": 0.5902255639097744,
"answer_agreement": 0.7619617224880383
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 613,
"failures_b": 625,
"failures_shared": 562,
"failure_jaccard": 0.8313609467455622,
"failure_recall_a_in_b": 0.9168026101141925,
"successes_a": 223,
"successes_b": 211,
"successes_shared": 160,
"success_jaccard": 0.583941605839416,
"answer_agreement": 0.7356459330143541
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 613,
"failures_b": 619,
"failures_shared": 561,
"failure_jaccard": 0.8360655737704918,
"failure_recall_a_in_b": 0.9151712887438825,
"successes_a": 223,
"successes_b": 217,
"successes_shared": 165,
"success_jaccard": 0.6,
"answer_agreement": 0.7452153110047847
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 613,
"failures_b": 614,
"failures_shared": 551,
"failure_jaccard": 0.8150887573964497,
"failure_recall_a_in_b": 0.8988580750407831,
"successes_a": 223,
"successes_b": 222,
"successes_shared": 160,
"success_jaccard": 0.5614035087719298,
"answer_agreement": 0.7260765550239234
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 631,
"failures_b": 636,
"failures_shared": 588,
"failure_jaccard": 0.865979381443299,
"failure_recall_a_in_b": 0.9318541996830428,
"successes_a": 205,
"successes_b": 200,
"successes_shared": 157,
"success_jaccard": 0.6330645161290323,
"answer_agreement": 0.7811004784688995
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 631,
"failures_b": 625,
"failures_shared": 580,
"failure_jaccard": 0.8579881656804734,
"failure_recall_a_in_b": 0.919175911251981,
"successes_a": 205,
"successes_b": 211,
"successes_shared": 160,
"success_jaccard": 0.625,
"answer_agreement": 0.7667464114832536
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 631,
"failures_b": 619,
"failures_shared": 571,
"failure_jaccard": 0.8409425625920471,
"failure_recall_a_in_b": 0.9049128367670365,
"successes_a": 205,
"successes_b": 217,
"successes_shared": 157,
"success_jaccard": 0.5924528301886792,
"answer_agreement": 0.7667464114832536
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 631,
"failures_b": 614,
"failures_shared": 569,
"failure_jaccard": 0.841715976331361,
"failure_recall_a_in_b": 0.901743264659271,
"successes_a": 205,
"successes_b": 222,
"successes_shared": 160,
"success_jaccard": 0.599250936329588,
"answer_agreement": 0.7464114832535885
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 636,
"failures_b": 625,
"failures_shared": 582,
"failure_jaccard": 0.8571428571428571,
"failure_recall_a_in_b": 0.9150943396226415,
"successes_a": 200,
"successes_b": 211,
"successes_shared": 157,
"success_jaccard": 0.6181102362204725,
"answer_agreement": 0.7739234449760766
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 636,
"failures_b": 619,
"failures_shared": 582,
"failure_jaccard": 0.8647845468053492,
"failure_recall_a_in_b": 0.9150943396226415,
"successes_a": 200,
"successes_b": 217,
"successes_shared": 163,
"success_jaccard": 0.6417322834645669,
"answer_agreement": 0.7918660287081339
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 636,
"failures_b": 614,
"failures_shared": 578,
"failure_jaccard": 0.8601190476190477,
"failure_recall_a_in_b": 0.9088050314465409,
"successes_a": 200,
"successes_b": 222,
"successes_shared": 164,
"success_jaccard": 0.6356589147286822,
"answer_agreement": 0.7715311004784688
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 625,
"failures_b": 619,
"failures_shared": 574,
"failure_jaccard": 0.8567164179104477,
"failure_recall_a_in_b": 0.9184,
"successes_a": 211,
"successes_b": 217,
"successes_shared": 166,
"success_jaccard": 0.6335877862595419,
"answer_agreement": 0.7799043062200957
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 625,
"failures_b": 614,
"failures_shared": 573,
"failure_jaccard": 0.8603603603603603,
"failure_recall_a_in_b": 0.9168,
"successes_a": 211,
"successes_b": 222,
"successes_shared": 170,
"success_jaccard": 0.6463878326996197,
"answer_agreement": 0.7811004784688995
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 619,
"failures_b": 614,
"failures_shared": 578,
"failure_jaccard": 0.8824427480916031,
"failure_recall_a_in_b": 0.9337641357027464,
"successes_a": 217,
"successes_b": 222,
"successes_shared": 181,
"success_jaccard": 0.7015503875968992,
"answer_agreement": 0.8014354066985646
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-7B-Instruct",
"task": "vsr",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 2195,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 2195,
"accuracy": 0.8223234624145785,
"num_failures": 390,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 2195,
"accuracy": 0.8077448747152619,
"num_failures": 422,
"flip_rate_vs_reference": 0.06013667425968109,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 2195,
"accuracy": 0.8027334851936219,
"num_failures": 433,
"flip_rate_vs_reference": 0.07334851936218678,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 2195,
"accuracy": 0.8068337129840547,
"num_failures": 424,
"flip_rate_vs_reference": 0.06560364464692484,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 2195,
"accuracy": 0.8022779043280183,
"num_failures": 434,
"flip_rate_vs_reference": 0.08109339407744874,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 2195,
"accuracy": 0.7881548974943052,
"num_failures": 465,
"flip_rate_vs_reference": 0.09339407744874716,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.034168564920273314,
"stability": {
"always_correct": 1579,
"always_wrong": 268,
"unstable": 348,
"always_correct_rate": 0.7193621867881549,
"always_wrong_rate": 0.12209567198177676,
"unstable_rate": 0.15854214123006835
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 390,
"failures_b": 422,
"failures_shared": 340,
"failure_jaccard": 0.7203389830508474,
"failure_recall_a_in_b": 0.8717948717948718,
"successes_a": 1805,
"successes_b": 1773,
"successes_shared": 1723,
"success_jaccard": 0.9288409703504044,
"answer_agreement": 0.939863325740319
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 390,
"failures_b": 433,
"failures_shared": 331,
"failure_jaccard": 0.6727642276422764,
"failure_recall_a_in_b": 0.8487179487179487,
"successes_a": 1805,
"successes_b": 1762,
"successes_shared": 1703,
"success_jaccard": 0.9136266094420601,
"answer_agreement": 0.9266514806378132
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 390,
"failures_b": 424,
"failures_shared": 335,
"failure_jaccard": 0.6993736951983298,
"failure_recall_a_in_b": 0.8589743589743589,
"successes_a": 1805,
"successes_b": 1771,
"successes_shared": 1716,
"success_jaccard": 0.9225806451612903,
"answer_agreement": 0.9343963553530752
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 390,
"failures_b": 434,
"failures_shared": 323,
"failure_jaccard": 0.6447105788423154,
"failure_recall_a_in_b": 0.8282051282051283,
"successes_a": 1805,
"successes_b": 1761,
"successes_shared": 1694,
"success_jaccard": 0.9049145299145299,
"answer_agreement": 0.9189066059225512
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 390,
"failures_b": 465,
"failures_shared": 325,
"failure_jaccard": 0.6132075471698113,
"failure_recall_a_in_b": 0.8333333333333334,
"successes_a": 1805,
"successes_b": 1730,
"successes_shared": 1665,
"success_jaccard": 0.8903743315508021,
"answer_agreement": 0.9066059225512528
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 422,
"failures_b": 433,
"failures_shared": 362,
"failure_jaccard": 0.7342799188640974,
"failure_recall_a_in_b": 0.8578199052132701,
"successes_a": 1773,
"successes_b": 1762,
"successes_shared": 1702,
"success_jaccard": 0.928532460447354,
"answer_agreement": 0.9403189066059225
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 422,
"failures_b": 424,
"failures_shared": 355,
"failure_jaccard": 0.7230142566191446,
"failure_recall_a_in_b": 0.8412322274881516,
"successes_a": 1773,
"successes_b": 1771,
"successes_shared": 1704,
"success_jaccard": 0.9260869565217391,
"answer_agreement": 0.9380410022779043
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 422,
"failures_b": 434,
"failures_shared": 345,
"failure_jaccard": 0.675146771037182,
"failure_recall_a_in_b": 0.8175355450236966,
"successes_a": 1773,
"successes_b": 1761,
"successes_shared": 1684,
"success_jaccard": 0.9102702702702703,
"answer_agreement": 0.924373576309795
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 422,
"failures_b": 465,
"failures_shared": 348,
"failure_jaccard": 0.6456400742115028,
"failure_recall_a_in_b": 0.8246445497630331,
"successes_a": 1773,
"successes_b": 1730,
"successes_shared": 1656,
"success_jaccard": 0.8965890633459664,
"answer_agreement": 0.9129840546697039
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 433,
"failures_b": 424,
"failures_shared": 359,
"failure_jaccard": 0.7208835341365462,
"failure_recall_a_in_b": 0.8290993071593533,
"successes_a": 1762,
"successes_b": 1771,
"successes_shared": 1697,
"success_jaccard": 0.9242919389978214,
"answer_agreement": 0.9366742596810934
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 433,
"failures_b": 434,
"failures_shared": 357,
"failure_jaccard": 0.7,
"failure_recall_a_in_b": 0.8244803695150116,
"successes_a": 1762,
"successes_b": 1761,
"successes_shared": 1685,
"success_jaccard": 0.9167573449401524,
"answer_agreement": 0.9302961275626423
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 433,
"failures_b": 465,
"failures_shared": 360,
"failure_jaccard": 0.6691449814126395,
"failure_recall_a_in_b": 0.8314087759815243,
"successes_a": 1762,
"successes_b": 1730,
"successes_shared": 1657,
"success_jaccard": 0.9029972752043597,
"answer_agreement": 0.9189066059225512
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 424,
"failures_b": 434,
"failures_shared": 361,
"failure_jaccard": 0.7263581488933601,
"failure_recall_a_in_b": 0.8514150943396226,
"successes_a": 1771,
"successes_b": 1761,
"successes_shared": 1698,
"success_jaccard": 0.925845147219193,
"answer_agreement": 0.9380410022779043
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 424,
"failures_b": 465,
"failures_shared": 363,
"failure_jaccard": 0.6901140684410646,
"failure_recall_a_in_b": 0.8561320754716981,
"successes_a": 1771,
"successes_b": 1730,
"successes_shared": 1669,
"success_jaccard": 0.9110262008733624,
"answer_agreement": 0.9257403189066059
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 434,
"failures_b": 465,
"failures_shared": 376,
"failure_jaccard": 0.7189292543021033,
"failure_recall_a_in_b": 0.8663594470046083,
"successes_a": 1761,
"successes_b": 1730,
"successes_shared": 1672,
"success_jaccard": 0.9191863661352392,
"answer_agreement": 0.9330296127562643
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-7B-Instruct",
"task": "youtube_level1",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 821,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 821,
"accuracy": 0.4470158343483557,
"num_failures": 454,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 821,
"accuracy": 0.4482338611449452,
"num_failures": 453,
"flip_rate_vs_reference": 0.1218026796589525,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 821,
"accuracy": 0.45554202192448234,
"num_failures": 447,
"flip_rate_vs_reference": 0.146163215590743,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 821,
"accuracy": 0.46650426309378806,
"num_failures": 438,
"flip_rate_vs_reference": 0.14981729598051158,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 821,
"accuracy": 0.44214372716199757,
"num_failures": 458,
"flip_rate_vs_reference": 0.17295980511571254,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 821,
"accuracy": 0.4518879415347138,
"num_failures": 450,
"flip_rate_vs_reference": 0.20219244823386115,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.024360535931790495,
"stability": {
"always_correct": 284,
"always_wrong": 366,
"unstable": 171,
"always_correct_rate": 0.3459196102314251,
"always_wrong_rate": 0.4457978075517661,
"unstable_rate": 0.20828258221680876
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 454,
"failures_b": 453,
"failures_shared": 417,
"failure_jaccard": 0.8510204081632653,
"failure_recall_a_in_b": 0.9185022026431718,
"successes_a": 367,
"successes_b": 368,
"successes_shared": 331,
"success_jaccard": 0.8193069306930693,
"answer_agreement": 0.8781973203410475
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 454,
"failures_b": 447,
"failures_shared": 413,
"failure_jaccard": 0.8463114754098361,
"failure_recall_a_in_b": 0.9096916299559471,
"successes_a": 367,
"successes_b": 374,
"successes_shared": 333,
"success_jaccard": 0.8161764705882353,
"answer_agreement": 0.853836784409257
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 454,
"failures_b": 438,
"failures_shared": 405,
"failure_jaccard": 0.8316221765913757,
"failure_recall_a_in_b": 0.8920704845814978,
"successes_a": 367,
"successes_b": 383,
"successes_shared": 334,
"success_jaccard": 0.8028846153846154,
"answer_agreement": 0.8501827040194885
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 454,
"failures_b": 458,
"failures_shared": 409,
"failure_jaccard": 0.8131212723658051,
"failure_recall_a_in_b": 0.9008810572687225,
"successes_a": 367,
"successes_b": 363,
"successes_shared": 318,
"success_jaccard": 0.7718446601941747,
"answer_agreement": 0.8270401948842875
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 454,
"failures_b": 450,
"failures_shared": 394,
"failure_jaccard": 0.7725490196078432,
"failure_recall_a_in_b": 0.8678414096916299,
"successes_a": 367,
"successes_b": 371,
"successes_shared": 311,
"success_jaccard": 0.7283372365339579,
"answer_agreement": 0.7978075517661388
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 453,
"failures_b": 447,
"failures_shared": 419,
"failure_jaccard": 0.8711018711018711,
"failure_recall_a_in_b": 0.9249448123620309,
"successes_a": 368,
"successes_b": 374,
"successes_shared": 340,
"success_jaccard": 0.845771144278607,
"answer_agreement": 0.8830694275274056
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 453,
"failures_b": 438,
"failures_shared": 412,
"failure_jaccard": 0.860125260960334,
"failure_recall_a_in_b": 0.9094922737306843,
"successes_a": 368,
"successes_b": 383,
"successes_shared": 342,
"success_jaccard": 0.8361858190709046,
"answer_agreement": 0.8842874543239951
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 453,
"failures_b": 458,
"failures_shared": 416,
"failure_jaccard": 0.8404040404040404,
"failure_recall_a_in_b": 0.9183222958057395,
"successes_a": 368,
"successes_b": 363,
"successes_shared": 326,
"success_jaccard": 0.8049382716049382,
"answer_agreement": 0.8623629719853837
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 453,
"failures_b": 450,
"failures_shared": 408,
"failure_jaccard": 0.8242424242424242,
"failure_recall_a_in_b": 0.9006622516556292,
"successes_a": 368,
"successes_b": 371,
"successes_shared": 326,
"success_jaccard": 0.7893462469733656,
"answer_agreement": 0.8428745432399513
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 447,
"failures_b": 438,
"failures_shared": 414,
"failure_jaccard": 0.8789808917197452,
"failure_recall_a_in_b": 0.9261744966442953,
"successes_a": 374,
"successes_b": 383,
"successes_shared": 350,
"success_jaccard": 0.85995085995086,
"answer_agreement": 0.8842874543239951
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 447,
"failures_b": 458,
"failures_shared": 420,
"failure_jaccard": 0.865979381443299,
"failure_recall_a_in_b": 0.9395973154362416,
"successes_a": 374,
"successes_b": 363,
"successes_shared": 336,
"success_jaccard": 0.8379052369077307,
"answer_agreement": 0.8721071863580999
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 447,
"failures_b": 450,
"failures_shared": 411,
"failure_jaccard": 0.845679012345679,
"failure_recall_a_in_b": 0.9194630872483222,
"successes_a": 374,
"successes_b": 371,
"successes_shared": 335,
"success_jaccard": 0.8170731707317073,
"answer_agreement": 0.8501827040194885
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 438,
"failures_b": 458,
"failures_shared": 416,
"failure_jaccard": 0.8666666666666667,
"failure_recall_a_in_b": 0.9497716894977168,
"successes_a": 383,
"successes_b": 363,
"successes_shared": 341,
"success_jaccard": 0.8419753086419753,
"answer_agreement": 0.8952496954933008
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 438,
"failures_b": 450,
"failures_shared": 406,
"failure_jaccard": 0.8423236514522822,
"failure_recall_a_in_b": 0.9269406392694064,
"successes_a": 383,
"successes_b": 371,
"successes_shared": 339,
"success_jaccard": 0.8168674698795181,
"answer_agreement": 0.8611449451887941
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 458,
"failures_b": 450,
"failures_shared": 423,
"failure_jaccard": 0.8721649484536083,
"failure_recall_a_in_b": 0.9235807860262009,
"successes_a": 363,
"successes_b": 371,
"successes_shared": 336,
"success_jaccard": 0.8442211055276382,
"answer_agreement": 0.879415347137637
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-7B-Instruct",
"task": "youtube_level2",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 764,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 764,
"accuracy": 0.4031413612565445,
"num_failures": 456,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 764,
"accuracy": 0.39397905759162305,
"num_failures": 463,
"flip_rate_vs_reference": 0.14397905759162305,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 764,
"accuracy": 0.39790575916230364,
"num_failures": 460,
"flip_rate_vs_reference": 0.13219895287958114,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 764,
"accuracy": 0.4018324607329843,
"num_failures": 457,
"flip_rate_vs_reference": 0.17277486910994763,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 764,
"accuracy": 0.3769633507853403,
"num_failures": 476,
"flip_rate_vs_reference": 0.18979057591623036,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 764,
"accuracy": 0.3769633507853403,
"num_failures": 476,
"flip_rate_vs_reference": 0.20026178010471204,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.02617801047120416,
"stability": {
"always_correct": 221,
"always_wrong": 382,
"unstable": 161,
"always_correct_rate": 0.2892670157068063,
"always_wrong_rate": 0.5,
"unstable_rate": 0.2107329842931937
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 456,
"failures_b": 463,
"failures_shared": 429,
"failure_jaccard": 0.8755102040816326,
"failure_recall_a_in_b": 0.9407894736842105,
"successes_a": 308,
"successes_b": 301,
"successes_shared": 274,
"success_jaccard": 0.817910447761194,
"answer_agreement": 0.856020942408377
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 456,
"failures_b": 460,
"failures_shared": 428,
"failure_jaccard": 0.8770491803278688,
"failure_recall_a_in_b": 0.9385964912280702,
"successes_a": 308,
"successes_b": 304,
"successes_shared": 276,
"success_jaccard": 0.8214285714285714,
"answer_agreement": 0.8678010471204188
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 456,
"failures_b": 457,
"failures_shared": 415,
"failure_jaccard": 0.8333333333333334,
"failure_recall_a_in_b": 0.9100877192982456,
"successes_a": 308,
"successes_b": 307,
"successes_shared": 266,
"success_jaccard": 0.7621776504297995,
"answer_agreement": 0.8272251308900523
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 456,
"failures_b": 476,
"failures_shared": 421,
"failure_jaccard": 0.8238747553816047,
"failure_recall_a_in_b": 0.9232456140350878,
"successes_a": 308,
"successes_b": 288,
"successes_shared": 253,
"success_jaccard": 0.7376093294460642,
"answer_agreement": 0.8102094240837696
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 456,
"failures_b": 476,
"failures_shared": 418,
"failure_jaccard": 0.8132295719844358,
"failure_recall_a_in_b": 0.9166666666666666,
"successes_a": 308,
"successes_b": 288,
"successes_shared": 250,
"success_jaccard": 0.7225433526011561,
"answer_agreement": 0.7997382198952879
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 463,
"failures_b": 460,
"failures_shared": 436,
"failure_jaccard": 0.8952772073921971,
"failure_recall_a_in_b": 0.9416846652267818,
"successes_a": 301,
"successes_b": 304,
"successes_shared": 277,
"success_jaccard": 0.8445121951219512,
"answer_agreement": 0.8835078534031413
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 463,
"failures_b": 457,
"failures_shared": 423,
"failure_jaccard": 0.8511066398390342,
"failure_recall_a_in_b": 0.9136069114470843,
"successes_a": 301,
"successes_b": 307,
"successes_shared": 267,
"success_jaccard": 0.782991202346041,
"answer_agreement": 0.8481675392670157
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 463,
"failures_b": 476,
"failures_shared": 439,
"failure_jaccard": 0.878,
"failure_recall_a_in_b": 0.9481641468682506,
"successes_a": 301,
"successes_b": 288,
"successes_shared": 264,
"success_jaccard": 0.8123076923076923,
"answer_agreement": 0.8586387434554974
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 463,
"failures_b": 476,
"failures_shared": 433,
"failure_jaccard": 0.8557312252964426,
"failure_recall_a_in_b": 0.9352051835853131,
"successes_a": 301,
"successes_b": 288,
"successes_shared": 258,
"success_jaccard": 0.7794561933534743,
"answer_agreement": 0.849476439790576
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 460,
"failures_b": 457,
"failures_shared": 426,
"failure_jaccard": 0.8676171079429735,
"failure_recall_a_in_b": 0.9260869565217391,
"successes_a": 304,
"successes_b": 307,
"successes_shared": 273,
"success_jaccard": 0.8076923076923077,
"answer_agreement": 0.8691099476439791
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 460,
"failures_b": 476,
"failures_shared": 439,
"failure_jaccard": 0.8832997987927566,
"failure_recall_a_in_b": 0.9543478260869566,
"successes_a": 304,
"successes_b": 288,
"successes_shared": 267,
"success_jaccard": 0.8215384615384616,
"answer_agreement": 0.8717277486910995
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 460,
"failures_b": 476,
"failures_shared": 434,
"failure_jaccard": 0.8645418326693227,
"failure_recall_a_in_b": 0.9434782608695652,
"successes_a": 304,
"successes_b": 288,
"successes_shared": 262,
"success_jaccard": 0.793939393939394,
"answer_agreement": 0.8455497382198953
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 457,
"failures_b": 476,
"failures_shared": 430,
"failure_jaccard": 0.8548707753479126,
"failure_recall_a_in_b": 0.9409190371991247,
"successes_a": 307,
"successes_b": 288,
"successes_shared": 261,
"success_jaccard": 0.781437125748503,
"answer_agreement": 0.8638743455497382
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 457,
"failures_b": 476,
"failures_shared": 427,
"failure_jaccard": 0.8438735177865613,
"failure_recall_a_in_b": 0.9343544857768052,
"successes_a": 307,
"successes_b": 288,
"successes_shared": 258,
"success_jaccard": 0.7655786350148368,
"answer_agreement": 0.8337696335078534
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 476,
"failures_b": 476,
"failures_shared": 451,
"failure_jaccard": 0.9001996007984032,
"failure_recall_a_in_b": 0.9474789915966386,
"successes_a": 288,
"successes_b": 288,
"successes_shared": 263,
"success_jaccard": 0.8402555910543131,
"answer_agreement": 0.8887434554973822
}
]
},
{
"model_id": "Qwen/Qwen2.5-VL-7B-Instruct",
"task": "youtube_level3",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 836,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 836,
"accuracy": 0.2799043062200957,
"num_failures": 602,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 836,
"accuracy": 0.28708133971291866,
"num_failures": 596,
"flip_rate_vs_reference": 0.19736842105263158,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 836,
"accuracy": 0.2679425837320574,
"num_failures": 612,
"flip_rate_vs_reference": 0.20454545454545456,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 836,
"accuracy": 0.2822966507177033,
"num_failures": 600,
"flip_rate_vs_reference": 0.2284688995215311,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 836,
"accuracy": 0.27631578947368424,
"num_failures": 605,
"flip_rate_vs_reference": 0.2452153110047847,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 836,
"accuracy": 0.27392344497607657,
"num_failures": 607,
"flip_rate_vs_reference": 0.23205741626794257,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.019138755980861233,
"stability": {
"always_correct": 143,
"always_wrong": 503,
"unstable": 190,
"always_correct_rate": 0.17105263157894737,
"always_wrong_rate": 0.6016746411483254,
"unstable_rate": 0.22727272727272727
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 602,
"failures_b": 596,
"failures_shared": 555,
"failure_jaccard": 0.8631415241057543,
"failure_recall_a_in_b": 0.9219269102990033,
"successes_a": 234,
"successes_b": 240,
"successes_shared": 193,
"success_jaccard": 0.6868327402135231,
"answer_agreement": 0.8026315789473685
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 602,
"failures_b": 612,
"failures_shared": 567,
"failure_jaccard": 0.8763523956723338,
"failure_recall_a_in_b": 0.9418604651162791,
"successes_a": 234,
"successes_b": 224,
"successes_shared": 189,
"success_jaccard": 0.7026022304832714,
"answer_agreement": 0.7954545454545454
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 602,
"failures_b": 600,
"failures_shared": 556,
"failure_jaccard": 0.8606811145510835,
"failure_recall_a_in_b": 0.9235880398671097,
"successes_a": 234,
"successes_b": 236,
"successes_shared": 190,
"success_jaccard": 0.6785714285714286,
"answer_agreement": 0.7715311004784688
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 602,
"failures_b": 605,
"failures_shared": 551,
"failure_jaccard": 0.8399390243902439,
"failure_recall_a_in_b": 0.915282392026578,
"successes_a": 234,
"successes_b": 231,
"successes_shared": 180,
"success_jaccard": 0.631578947368421,
"answer_agreement": 0.7547846889952153
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 602,
"failures_b": 607,
"failures_shared": 556,
"failure_jaccard": 0.8514548238897397,
"failure_recall_a_in_b": 0.9235880398671097,
"successes_a": 234,
"successes_b": 229,
"successes_shared": 183,
"success_jaccard": 0.6535714285714286,
"answer_agreement": 0.7679425837320574
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 596,
"failures_b": 612,
"failures_shared": 561,
"failure_jaccard": 0.8670788253477589,
"failure_recall_a_in_b": 0.9412751677852349,
"successes_a": 240,
"successes_b": 224,
"successes_shared": 189,
"success_jaccard": 0.6872727272727273,
"answer_agreement": 0.8205741626794258
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 596,
"failures_b": 600,
"failures_shared": 559,
"failure_jaccard": 0.8775510204081632,
"failure_recall_a_in_b": 0.9379194630872483,
"successes_a": 240,
"successes_b": 236,
"successes_shared": 199,
"success_jaccard": 0.7184115523465704,
"answer_agreement": 0.8098086124401914
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 596,
"failures_b": 605,
"failures_shared": 554,
"failure_jaccard": 0.8562596599690881,
"failure_recall_a_in_b": 0.9295302013422819,
"successes_a": 240,
"successes_b": 231,
"successes_shared": 189,
"success_jaccard": 0.6702127659574468,
"answer_agreement": 0.7954545454545454
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 596,
"failures_b": 607,
"failures_shared": 552,
"failure_jaccard": 0.847926267281106,
"failure_recall_a_in_b": 0.9261744966442953,
"successes_a": 240,
"successes_b": 229,
"successes_shared": 185,
"success_jaccard": 0.6514084507042254,
"answer_agreement": 0.7715311004784688
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 612,
"failures_b": 600,
"failures_shared": 565,
"failure_jaccard": 0.8732612055641422,
"failure_recall_a_in_b": 0.923202614379085,
"successes_a": 224,
"successes_b": 236,
"successes_shared": 189,
"success_jaccard": 0.6974169741697417,
"answer_agreement": 0.8169856459330144
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 612,
"failures_b": 605,
"failures_shared": 566,
"failure_jaccard": 0.869431643625192,
"failure_recall_a_in_b": 0.9248366013071896,
"successes_a": 224,
"successes_b": 231,
"successes_shared": 185,
"success_jaccard": 0.6851851851851852,
"answer_agreement": 0.8169856459330144
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 612,
"failures_b": 607,
"failures_shared": 566,
"failure_jaccard": 0.8667687595712098,
"failure_recall_a_in_b": 0.9248366013071896,
"successes_a": 224,
"successes_b": 229,
"successes_shared": 183,
"success_jaccard": 0.6777777777777778,
"answer_agreement": 0.8002392344497608
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 600,
"failures_b": 605,
"failures_shared": 565,
"failure_jaccard": 0.8828125,
"failure_recall_a_in_b": 0.9416666666666667,
"successes_a": 236,
"successes_b": 231,
"successes_shared": 196,
"success_jaccard": 0.7232472324723247,
"answer_agreement": 0.8217703349282297
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 600,
"failures_b": 607,
"failures_shared": 560,
"failure_jaccard": 0.865533230293663,
"failure_recall_a_in_b": 0.9333333333333333,
"successes_a": 236,
"successes_b": 229,
"successes_shared": 189,
"success_jaccard": 0.6847826086956522,
"answer_agreement": 0.7942583732057417
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 605,
"failures_b": 607,
"failures_shared": 564,
"failure_jaccard": 0.8703703703703703,
"failure_recall_a_in_b": 0.9322314049586777,
"successes_a": 231,
"successes_b": 229,
"successes_shared": 188,
"success_jaccard": 0.6911764705882353,
"answer_agreement": 0.8014354066985646
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B",
"task": "vsr",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 2195,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 2195,
"accuracy": 0.8214123006833713,
"num_failures": 392,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 2195,
"accuracy": 0.8173120728929385,
"num_failures": 401,
"flip_rate_vs_reference": 0.04510250569476082,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 2195,
"accuracy": 0.8186788154897494,
"num_failures": 398,
"flip_rate_vs_reference": 0.04738041002277904,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 2195,
"accuracy": 0.8113895216400911,
"num_failures": 414,
"flip_rate_vs_reference": 0.06287015945330296,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 2195,
"accuracy": 0.8072892938496583,
"num_failures": 423,
"flip_rate_vs_reference": 0.07425968109339408,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 2195,
"accuracy": 0.8045558086560365,
"num_failures": 429,
"flip_rate_vs_reference": 0.07608200455580866,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.016856492027334857,
"stability": {
"always_correct": 1635,
"always_wrong": 283,
"unstable": 277,
"always_correct_rate": 0.744874715261959,
"always_wrong_rate": 0.12892938496583142,
"unstable_rate": 0.12619589977220957
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 392,
"failures_b": 401,
"failures_shared": 347,
"failure_jaccard": 0.7780269058295964,
"failure_recall_a_in_b": 0.8852040816326531,
"successes_a": 1803,
"successes_b": 1794,
"successes_shared": 1749,
"success_jaccard": 0.9464285714285714,
"answer_agreement": 0.9548974943052392
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 392,
"failures_b": 398,
"failures_shared": 343,
"failure_jaccard": 0.767337807606264,
"failure_recall_a_in_b": 0.875,
"successes_a": 1803,
"successes_b": 1797,
"successes_shared": 1748,
"success_jaccard": 0.9438444924406048,
"answer_agreement": 0.952619589977221
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 392,
"failures_b": 414,
"failures_shared": 334,
"failure_jaccard": 0.7076271186440678,
"failure_recall_a_in_b": 0.8520408163265306,
"successes_a": 1803,
"successes_b": 1781,
"successes_shared": 1723,
"success_jaccard": 0.9258463191832348,
"answer_agreement": 0.937129840546697
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 392,
"failures_b": 423,
"failures_shared": 326,
"failure_jaccard": 0.6666666666666666,
"failure_recall_a_in_b": 0.8316326530612245,
"successes_a": 1803,
"successes_b": 1772,
"successes_shared": 1706,
"success_jaccard": 0.9127875869448903,
"answer_agreement": 0.9257403189066059
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 392,
"failures_b": 429,
"failures_shared": 327,
"failure_jaccard": 0.6619433198380567,
"failure_recall_a_in_b": 0.8341836734693877,
"successes_a": 1803,
"successes_b": 1766,
"successes_shared": 1701,
"success_jaccard": 0.9105995717344754,
"answer_agreement": 0.9239179954441914
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 401,
"failures_b": 398,
"failures_shared": 356,
"failure_jaccard": 0.8036117381489842,
"failure_recall_a_in_b": 0.8877805486284289,
"successes_a": 1794,
"successes_b": 1797,
"successes_shared": 1752,
"success_jaccard": 0.9526916802610114,
"answer_agreement": 0.9603644646924829
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 401,
"failures_b": 414,
"failures_shared": 350,
"failure_jaccard": 0.7526881720430108,
"failure_recall_a_in_b": 0.8728179551122195,
"successes_a": 1794,
"successes_b": 1781,
"successes_shared": 1730,
"success_jaccard": 0.9376693766937669,
"answer_agreement": 0.9476082004555809
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 401,
"failures_b": 423,
"failures_shared": 341,
"failure_jaccard": 0.7060041407867494,
"failure_recall_a_in_b": 0.8503740648379052,
"successes_a": 1794,
"successes_b": 1772,
"successes_shared": 1712,
"success_jaccard": 0.9234088457389428,
"answer_agreement": 0.9353075170842825
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 401,
"failures_b": 429,
"failures_shared": 337,
"failure_jaccard": 0.6835699797160243,
"failure_recall_a_in_b": 0.8403990024937655,
"successes_a": 1794,
"successes_b": 1766,
"successes_shared": 1702,
"success_jaccard": 0.9160387513455328,
"answer_agreement": 0.9289293849658314
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 398,
"failures_b": 414,
"failures_shared": 355,
"failure_jaccard": 0.7768052516411379,
"failure_recall_a_in_b": 0.8919597989949749,
"successes_a": 1797,
"successes_b": 1781,
"successes_shared": 1738,
"success_jaccard": 0.9445652173913044,
"answer_agreement": 0.9535307517084283
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 398,
"failures_b": 423,
"failures_shared": 350,
"failure_jaccard": 0.7430997876857749,
"failure_recall_a_in_b": 0.8793969849246231,
"successes_a": 1797,
"successes_b": 1772,
"successes_shared": 1724,
"success_jaccard": 0.9344173441734417,
"answer_agreement": 0.944874715261959
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 398,
"failures_b": 429,
"failures_shared": 341,
"failure_jaccard": 0.7016460905349794,
"failure_recall_a_in_b": 0.8567839195979899,
"successes_a": 1797,
"successes_b": 1766,
"successes_shared": 1709,
"success_jaccard": 0.9217907227615966,
"answer_agreement": 0.9339407744874715
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 414,
"failures_b": 423,
"failures_shared": 359,
"failure_jaccard": 0.7510460251046025,
"failure_recall_a_in_b": 0.8671497584541062,
"successes_a": 1781,
"successes_b": 1772,
"successes_shared": 1717,
"success_jaccard": 0.9351851851851852,
"answer_agreement": 0.9457858769931663
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 414,
"failures_b": 429,
"failures_shared": 354,
"failure_jaccard": 0.7239263803680982,
"failure_recall_a_in_b": 0.855072463768116,
"successes_a": 1781,
"successes_b": 1766,
"successes_shared": 1706,
"success_jaccard": 0.9266702878870179,
"answer_agreement": 0.9384965831435079
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 423,
"failures_b": 429,
"failures_shared": 372,
"failure_jaccard": 0.775,
"failure_recall_a_in_b": 0.8794326241134752,
"successes_a": 1772,
"successes_b": 1766,
"successes_shared": 1715,
"success_jaccard": 0.9407569939659901,
"answer_agreement": 0.9507972665148063
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B",
"task": "youtube_level1",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 821,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 821,
"accuracy": 0.5274056029232643,
"num_failures": 388,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0036540803897685747
},
{
"baseline": "0.05",
"num_samples": 821,
"accuracy": 0.5213154689403167,
"num_failures": 393,
"flip_rate_vs_reference": 0.10353227771010962,
"unparsed_rate": 0.007308160779537149
},
{
"baseline": "0.10",
"num_samples": 821,
"accuracy": 0.5188794153471377,
"num_failures": 395,
"flip_rate_vs_reference": 0.12545676004872108,
"unparsed_rate": 0.001218026796589525
},
{
"baseline": "0.15",
"num_samples": 821,
"accuracy": 0.5408038976857491,
"num_failures": 377,
"flip_rate_vs_reference": 0.14007308160779536,
"unparsed_rate": 0.0048721071863581
},
{
"baseline": "0.20",
"num_samples": 821,
"accuracy": 0.5347137637028014,
"num_failures": 382,
"flip_rate_vs_reference": 0.15956151035322777,
"unparsed_rate": 0.001218026796589525
},
{
"baseline": "0.25",
"num_samples": 821,
"accuracy": 0.5347137637028014,
"num_failures": 382,
"flip_rate_vs_reference": 0.1656516443361754,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.02192448233861144,
"stability": {
"always_correct": 343,
"always_wrong": 296,
"unstable": 182,
"always_correct_rate": 0.4177831912302071,
"always_wrong_rate": 0.36053593179049936,
"unstable_rate": 0.22168087697929353
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 388,
"failures_b": 393,
"failures_shared": 361,
"failure_jaccard": 0.8595238095238096,
"failure_recall_a_in_b": 0.9304123711340206,
"successes_a": 433,
"successes_b": 428,
"successes_shared": 401,
"success_jaccard": 0.8717391304347826,
"answer_agreement": 0.8964677222898904
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 388,
"failures_b": 395,
"failures_shared": 354,
"failure_jaccard": 0.8251748251748252,
"failure_recall_a_in_b": 0.9123711340206185,
"successes_a": 433,
"successes_b": 426,
"successes_shared": 392,
"success_jaccard": 0.8394004282655246,
"answer_agreement": 0.8745432399512789
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 388,
"failures_b": 377,
"failures_shared": 342,
"failure_jaccard": 0.8085106382978723,
"failure_recall_a_in_b": 0.8814432989690721,
"successes_a": 433,
"successes_b": 444,
"successes_shared": 398,
"success_jaccard": 0.8308977035490606,
"answer_agreement": 0.8599269183922046
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 388,
"failures_b": 382,
"failures_shared": 339,
"failure_jaccard": 0.7865429234338747,
"failure_recall_a_in_b": 0.8737113402061856,
"successes_a": 433,
"successes_b": 439,
"successes_shared": 390,
"success_jaccard": 0.8091286307053942,
"answer_agreement": 0.8404384896467723
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 388,
"failures_b": 382,
"failures_shared": 338,
"failure_jaccard": 0.7824074074074074,
"failure_recall_a_in_b": 0.8711340206185567,
"successes_a": 433,
"successes_b": 439,
"successes_shared": 389,
"success_jaccard": 0.8053830227743272,
"answer_agreement": 0.8343483556638246
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 393,
"failures_b": 395,
"failures_shared": 363,
"failure_jaccard": 0.8541176470588235,
"failure_recall_a_in_b": 0.9236641221374046,
"successes_a": 428,
"successes_b": 426,
"successes_shared": 396,
"success_jaccard": 0.8646288209606987,
"answer_agreement": 0.8855054811205847
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 393,
"failures_b": 377,
"failures_shared": 342,
"failure_jaccard": 0.7990654205607477,
"failure_recall_a_in_b": 0.8702290076335878,
"successes_a": 428,
"successes_b": 444,
"successes_shared": 393,
"success_jaccard": 0.8204592901878914,
"answer_agreement": 0.8635809987819733
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 393,
"failures_b": 382,
"failures_shared": 343,
"failure_jaccard": 0.7939814814814815,
"failure_recall_a_in_b": 0.8727735368956743,
"successes_a": 428,
"successes_b": 439,
"successes_shared": 389,
"success_jaccard": 0.8138075313807531,
"answer_agreement": 0.8453105968331304
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 393,
"failures_b": 382,
"failures_shared": 338,
"failure_jaccard": 0.7734553775743707,
"failure_recall_a_in_b": 0.8600508905852418,
"successes_a": 428,
"successes_b": 439,
"successes_shared": 384,
"success_jaccard": 0.7950310559006211,
"answer_agreement": 0.8258221680876979
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 395,
"failures_b": 377,
"failures_shared": 349,
"failure_jaccard": 0.8250591016548463,
"failure_recall_a_in_b": 0.8835443037974684,
"successes_a": 426,
"successes_b": 444,
"successes_shared": 398,
"success_jaccard": 0.8432203389830508,
"answer_agreement": 0.8781973203410475
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 395,
"failures_b": 382,
"failures_shared": 350,
"failure_jaccard": 0.819672131147541,
"failure_recall_a_in_b": 0.8860759493670886,
"successes_a": 426,
"successes_b": 439,
"successes_shared": 394,
"success_jaccard": 0.8365180467091295,
"answer_agreement": 0.8684531059683313
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 395,
"failures_b": 382,
"failures_shared": 347,
"failure_jaccard": 0.8069767441860465,
"failure_recall_a_in_b": 0.8784810126582279,
"successes_a": 426,
"successes_b": 439,
"successes_shared": 391,
"success_jaccard": 0.8248945147679325,
"answer_agreement": 0.8514007308160779
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 377,
"failures_b": 382,
"failures_shared": 341,
"failure_jaccard": 0.8157894736842105,
"failure_recall_a_in_b": 0.9045092838196287,
"successes_a": 444,
"successes_b": 439,
"successes_shared": 403,
"success_jaccard": 0.8395833333333333,
"answer_agreement": 0.8721071863580999
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 377,
"failures_b": 382,
"failures_shared": 338,
"failure_jaccard": 0.8028503562945368,
"failure_recall_a_in_b": 0.896551724137931,
"successes_a": 444,
"successes_b": 439,
"successes_shared": 400,
"success_jaccard": 0.8281573498964804,
"answer_agreement": 0.8611449451887941
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 382,
"failures_b": 382,
"failures_shared": 346,
"failure_jaccard": 0.8277511961722488,
"failure_recall_a_in_b": 0.9057591623036649,
"successes_a": 439,
"successes_b": 439,
"successes_shared": 403,
"success_jaccard": 0.848421052631579,
"answer_agreement": 0.8781973203410475
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B",
"task": "youtube_level2",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 764,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 764,
"accuracy": 0.3913612565445026,
"num_failures": 465,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 764,
"accuracy": 0.3900523560209424,
"num_failures": 466,
"flip_rate_vs_reference": 0.12172774869109948,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 764,
"accuracy": 0.38219895287958117,
"num_failures": 472,
"flip_rate_vs_reference": 0.14136125654450263,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 764,
"accuracy": 0.387434554973822,
"num_failures": 468,
"flip_rate_vs_reference": 0.16361256544502617,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 764,
"accuracy": 0.3795811518324607,
"num_failures": 474,
"flip_rate_vs_reference": 0.17146596858638743,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 764,
"accuracy": 0.38481675392670156,
"num_failures": 470,
"flip_rate_vs_reference": 0.18717277486910994,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.011780104712041883,
"stability": {
"always_correct": 230,
"always_wrong": 401,
"unstable": 133,
"always_correct_rate": 0.3010471204188482,
"always_wrong_rate": 0.524869109947644,
"unstable_rate": 0.17408376963350786
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 465,
"failures_b": 466,
"failures_shared": 438,
"failure_jaccard": 0.8884381338742393,
"failure_recall_a_in_b": 0.9419354838709677,
"successes_a": 299,
"successes_b": 298,
"successes_shared": 271,
"success_jaccard": 0.8312883435582822,
"answer_agreement": 0.8782722513089005
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 465,
"failures_b": 472,
"failures_shared": 438,
"failure_jaccard": 0.8777555110220441,
"failure_recall_a_in_b": 0.9419354838709677,
"successes_a": 299,
"successes_b": 292,
"successes_shared": 265,
"success_jaccard": 0.8128834355828221,
"answer_agreement": 0.8586387434554974
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 465,
"failures_b": 468,
"failures_shared": 433,
"failure_jaccard": 0.866,
"failure_recall_a_in_b": 0.9311827956989247,
"successes_a": 299,
"successes_b": 296,
"successes_shared": 264,
"success_jaccard": 0.797583081570997,
"answer_agreement": 0.8363874345549738
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 465,
"failures_b": 474,
"failures_shared": 434,
"failure_jaccard": 0.8594059405940594,
"failure_recall_a_in_b": 0.9333333333333333,
"successes_a": 299,
"successes_b": 290,
"successes_shared": 259,
"success_jaccard": 0.7848484848484848,
"answer_agreement": 0.8285340314136126
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 465,
"failures_b": 470,
"failures_shared": 423,
"failure_jaccard": 0.826171875,
"failure_recall_a_in_b": 0.9096774193548387,
"successes_a": 299,
"successes_b": 294,
"successes_shared": 252,
"success_jaccard": 0.7390029325513197,
"answer_agreement": 0.81282722513089
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 466,
"failures_b": 472,
"failures_shared": 441,
"failure_jaccard": 0.8873239436619719,
"failure_recall_a_in_b": 0.9463519313304721,
"successes_a": 298,
"successes_b": 292,
"successes_shared": 267,
"success_jaccard": 0.826625386996904,
"answer_agreement": 0.8821989528795812
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 466,
"failures_b": 468,
"failures_shared": 438,
"failure_jaccard": 0.8830645161290323,
"failure_recall_a_in_b": 0.9399141630901288,
"successes_a": 298,
"successes_b": 296,
"successes_shared": 268,
"success_jaccard": 0.8220858895705522,
"answer_agreement": 0.8664921465968587
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 466,
"failures_b": 474,
"failures_shared": 437,
"failure_jaccard": 0.8687872763419483,
"failure_recall_a_in_b": 0.9377682403433476,
"successes_a": 298,
"successes_b": 290,
"successes_shared": 261,
"success_jaccard": 0.7981651376146789,
"answer_agreement": 0.856020942408377
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 466,
"failures_b": 470,
"failures_shared": 432,
"failure_jaccard": 0.8571428571428571,
"failure_recall_a_in_b": 0.927038626609442,
"successes_a": 298,
"successes_b": 294,
"successes_shared": 260,
"success_jaccard": 0.7831325301204819,
"answer_agreement": 0.8599476439790575
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 472,
"failures_b": 468,
"failures_shared": 445,
"failure_jaccard": 0.898989898989899,
"failure_recall_a_in_b": 0.9427966101694916,
"successes_a": 292,
"successes_b": 296,
"successes_shared": 269,
"success_jaccard": 0.8432601880877743,
"answer_agreement": 0.8821989528795812
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 472,
"failures_b": 474,
"failures_shared": 447,
"failure_jaccard": 0.8957915831663327,
"failure_recall_a_in_b": 0.9470338983050848,
"successes_a": 292,
"successes_b": 290,
"successes_shared": 265,
"success_jaccard": 0.8359621451104101,
"answer_agreement": 0.8678010471204188
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 472,
"failures_b": 470,
"failures_shared": 438,
"failure_jaccard": 0.8690476190476191,
"failure_recall_a_in_b": 0.9279661016949152,
"successes_a": 292,
"successes_b": 294,
"successes_shared": 260,
"success_jaccard": 0.7975460122699386,
"answer_agreement": 0.856020942408377
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 468,
"failures_b": 474,
"failures_shared": 446,
"failure_jaccard": 0.8991935483870968,
"failure_recall_a_in_b": 0.9529914529914529,
"successes_a": 296,
"successes_b": 290,
"successes_shared": 268,
"success_jaccard": 0.8427672955974843,
"answer_agreement": 0.8704188481675392
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 468,
"failures_b": 470,
"failures_shared": 440,
"failure_jaccard": 0.8835341365461847,
"failure_recall_a_in_b": 0.9401709401709402,
"successes_a": 296,
"successes_b": 294,
"successes_shared": 266,
"success_jaccard": 0.8209876543209876,
"answer_agreement": 0.8704188481675392
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 474,
"failures_b": 470,
"failures_shared": 443,
"failure_jaccard": 0.8842315369261478,
"failure_recall_a_in_b": 0.9345991561181435,
"successes_a": 290,
"successes_b": 294,
"successes_shared": 263,
"success_jaccard": 0.8193146417445483,
"answer_agreement": 0.8520942408376964
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B",
"task": "youtube_level3",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 836,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 836,
"accuracy": 0.2930622009569378,
"num_failures": 591,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 836,
"accuracy": 0.29904306220095694,
"num_failures": 586,
"flip_rate_vs_reference": 0.23803827751196172,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 836,
"accuracy": 0.2966507177033493,
"num_failures": 588,
"flip_rate_vs_reference": 0.2595693779904306,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 836,
"accuracy": 0.3038277511961722,
"num_failures": 582,
"flip_rate_vs_reference": 0.2595693779904306,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 836,
"accuracy": 0.28827751196172247,
"num_failures": 595,
"flip_rate_vs_reference": 0.27870813397129185,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 836,
"accuracy": 0.2930622009569378,
"num_failures": 591,
"flip_rate_vs_reference": 0.29545454545454547,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.015550239234449759,
"stability": {
"always_correct": 144,
"always_wrong": 474,
"unstable": 218,
"always_correct_rate": 0.1722488038277512,
"always_wrong_rate": 0.5669856459330144,
"unstable_rate": 0.2607655502392344
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 591,
"failures_b": 586,
"failures_shared": 539,
"failure_jaccard": 0.8448275862068966,
"failure_recall_a_in_b": 0.9120135363790186,
"successes_a": 245,
"successes_b": 250,
"successes_shared": 198,
"success_jaccard": 0.6666666666666666,
"answer_agreement": 0.7619617224880383
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 591,
"failures_b": 588,
"failures_shared": 534,
"failure_jaccard": 0.827906976744186,
"failure_recall_a_in_b": 0.9035532994923858,
"successes_a": 245,
"successes_b": 248,
"successes_shared": 191,
"success_jaccard": 0.6324503311258278,
"answer_agreement": 0.7404306220095693
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 591,
"failures_b": 582,
"failures_shared": 530,
"failure_jaccard": 0.8242612752721618,
"failure_recall_a_in_b": 0.8967851099830795,
"successes_a": 245,
"successes_b": 254,
"successes_shared": 193,
"success_jaccard": 0.630718954248366,
"answer_agreement": 0.7404306220095693
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 591,
"failures_b": 595,
"failures_shared": 531,
"failure_jaccard": 0.8106870229007633,
"failure_recall_a_in_b": 0.8984771573604061,
"successes_a": 245,
"successes_b": 241,
"successes_shared": 181,
"success_jaccard": 0.5934426229508196,
"answer_agreement": 0.7212918660287081
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 591,
"failures_b": 591,
"failures_shared": 525,
"failure_jaccard": 0.7990867579908676,
"failure_recall_a_in_b": 0.8883248730964467,
"successes_a": 245,
"successes_b": 245,
"successes_shared": 179,
"success_jaccard": 0.5755627009646302,
"answer_agreement": 0.7045454545454546
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 586,
"failures_b": 588,
"failures_shared": 546,
"failure_jaccard": 0.8694267515923567,
"failure_recall_a_in_b": 0.931740614334471,
"successes_a": 250,
"successes_b": 248,
"successes_shared": 208,
"success_jaccard": 0.7172413793103448,
"answer_agreement": 0.8229665071770335
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 586,
"failures_b": 582,
"failures_shared": 540,
"failure_jaccard": 0.8598726114649682,
"failure_recall_a_in_b": 0.9215017064846417,
"successes_a": 250,
"successes_b": 254,
"successes_shared": 208,
"success_jaccard": 0.7027027027027027,
"answer_agreement": 0.7918660287081339
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 586,
"failures_b": 595,
"failures_shared": 546,
"failure_jaccard": 0.8598425196850393,
"failure_recall_a_in_b": 0.931740614334471,
"successes_a": 250,
"successes_b": 241,
"successes_shared": 201,
"success_jaccard": 0.6931034482758621,
"answer_agreement": 0.7811004784688995
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 586,
"failures_b": 591,
"failures_shared": 541,
"failure_jaccard": 0.85062893081761,
"failure_recall_a_in_b": 0.9232081911262798,
"successes_a": 250,
"successes_b": 245,
"successes_shared": 200,
"success_jaccard": 0.6779661016949152,
"answer_agreement": 0.7822966507177034
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 588,
"failures_b": 582,
"failures_shared": 544,
"failure_jaccard": 0.8690095846645367,
"failure_recall_a_in_b": 0.9251700680272109,
"successes_a": 248,
"successes_b": 254,
"successes_shared": 210,
"success_jaccard": 0.7191780821917808,
"answer_agreement": 0.8145933014354066
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 588,
"failures_b": 595,
"failures_shared": 555,
"failure_jaccard": 0.8837579617834395,
"failure_recall_a_in_b": 0.9438775510204082,
"successes_a": 248,
"successes_b": 241,
"successes_shared": 208,
"success_jaccard": 0.7402135231316725,
"answer_agreement": 0.8050239234449761
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 588,
"failures_b": 591,
"failures_shared": 538,
"failure_jaccard": 0.8393135725429017,
"failure_recall_a_in_b": 0.9149659863945578,
"successes_a": 248,
"successes_b": 245,
"successes_shared": 195,
"success_jaccard": 0.6543624161073825,
"answer_agreement": 0.7870813397129187
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 582,
"failures_b": 595,
"failures_shared": 551,
"failure_jaccard": 0.8801916932907349,
"failure_recall_a_in_b": 0.9467353951890034,
"successes_a": 254,
"successes_b": 241,
"successes_shared": 210,
"success_jaccard": 0.7368421052631579,
"answer_agreement": 0.819377990430622
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 582,
"failures_b": 591,
"failures_shared": 539,
"failure_jaccard": 0.8501577287066246,
"failure_recall_a_in_b": 0.9261168384879725,
"successes_a": 254,
"successes_b": 245,
"successes_shared": 202,
"success_jaccard": 0.6801346801346801,
"answer_agreement": 0.7930622009569378
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 595,
"failures_b": 591,
"failures_shared": 550,
"failure_jaccard": 0.8647798742138365,
"failure_recall_a_in_b": 0.9243697478991597,
"successes_a": 241,
"successes_b": 245,
"successes_shared": 200,
"success_jaccard": 0.6993006993006993,
"answer_agreement": 0.7918660287081339
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B-Base",
"task": "vsr",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 2195,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 2195,
"accuracy": 0.8318906605922551,
"num_failures": 369,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 2195,
"accuracy": 0.8314350797266514,
"num_failures": 370,
"flip_rate_vs_reference": 0.04601366742596811,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 2195,
"accuracy": 0.8323462414578587,
"num_failures": 368,
"flip_rate_vs_reference": 0.05330296127562642,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 2195,
"accuracy": 0.8277904328018223,
"num_failures": 378,
"flip_rate_vs_reference": 0.05968109339407745,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 2195,
"accuracy": 0.8241457858769932,
"num_failures": 386,
"flip_rate_vs_reference": 0.07881548974943052,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 2195,
"accuracy": 0.8164009111617312,
"num_failures": 403,
"flip_rate_vs_reference": 0.0856492027334852,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.015945330296127547,
"stability": {
"always_correct": 1653,
"always_wrong": 240,
"unstable": 302,
"always_correct_rate": 0.7530751708428246,
"always_wrong_rate": 0.10933940774487472,
"unstable_rate": 0.13758542141230068
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 369,
"failures_b": 370,
"failures_shared": 319,
"failure_jaccard": 0.7595238095238095,
"failure_recall_a_in_b": 0.8644986449864499,
"successes_a": 1826,
"successes_b": 1825,
"successes_shared": 1775,
"success_jaccard": 0.9461620469083155,
"answer_agreement": 0.9539863325740319
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 369,
"failures_b": 368,
"failures_shared": 310,
"failure_jaccard": 0.7259953161592506,
"failure_recall_a_in_b": 0.8401084010840109,
"successes_a": 1826,
"successes_b": 1827,
"successes_shared": 1768,
"success_jaccard": 0.9379310344827586,
"answer_agreement": 0.9466970387243736
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 369,
"failures_b": 378,
"failures_shared": 308,
"failure_jaccard": 0.7015945330296127,
"failure_recall_a_in_b": 0.8346883468834688,
"successes_a": 1826,
"successes_b": 1817,
"successes_shared": 1756,
"success_jaccard": 0.9305776364599894,
"answer_agreement": 0.9403189066059225
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 369,
"failures_b": 386,
"failures_shared": 291,
"failure_jaccard": 0.6271551724137931,
"failure_recall_a_in_b": 0.7886178861788617,
"successes_a": 1826,
"successes_b": 1809,
"successes_shared": 1731,
"success_jaccard": 0.9091386554621849,
"answer_agreement": 0.9211845102505695
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 369,
"failures_b": 403,
"failures_shared": 292,
"failure_jaccard": 0.6083333333333333,
"failure_recall_a_in_b": 0.7913279132791328,
"successes_a": 1826,
"successes_b": 1792,
"successes_shared": 1715,
"success_jaccard": 0.9012086179716238,
"answer_agreement": 0.9143507972665148
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 370,
"failures_b": 368,
"failures_shared": 319,
"failure_jaccard": 0.7613365155131265,
"failure_recall_a_in_b": 0.8621621621621621,
"successes_a": 1825,
"successes_b": 1827,
"successes_shared": 1776,
"success_jaccard": 0.9466950959488273,
"answer_agreement": 0.9544419134396356
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 370,
"failures_b": 378,
"failures_shared": 310,
"failure_jaccard": 0.7077625570776256,
"failure_recall_a_in_b": 0.8378378378378378,
"successes_a": 1825,
"successes_b": 1817,
"successes_shared": 1757,
"success_jaccard": 0.9320954907161804,
"answer_agreement": 0.9416856492027335
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 370,
"failures_b": 386,
"failures_shared": 302,
"failure_jaccard": 0.6651982378854625,
"failure_recall_a_in_b": 0.8162162162162162,
"successes_a": 1825,
"successes_b": 1809,
"successes_shared": 1741,
"success_jaccard": 0.9197041732699419,
"answer_agreement": 0.9307517084282461
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 370,
"failures_b": 403,
"failures_shared": 298,
"failure_jaccard": 0.6273684210526316,
"failure_recall_a_in_b": 0.8054054054054054,
"successes_a": 1825,
"successes_b": 1792,
"successes_shared": 1720,
"success_jaccard": 0.9066947812335266,
"answer_agreement": 0.9193621867881548
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 368,
"failures_b": 378,
"failures_shared": 310,
"failure_jaccard": 0.7110091743119266,
"failure_recall_a_in_b": 0.842391304347826,
"successes_a": 1827,
"successes_b": 1817,
"successes_shared": 1759,
"success_jaccard": 0.9331564986737401,
"answer_agreement": 0.9425968109339408
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 368,
"failures_b": 386,
"failures_shared": 316,
"failure_jaccard": 0.7214611872146118,
"failure_recall_a_in_b": 0.8586956521739131,
"successes_a": 1827,
"successes_b": 1809,
"successes_shared": 1757,
"success_jaccard": 0.9350718467269824,
"answer_agreement": 0.9444191343963554
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 368,
"failures_b": 403,
"failures_shared": 310,
"failure_jaccard": 0.6724511930585684,
"failure_recall_a_in_b": 0.842391304347826,
"successes_a": 1827,
"successes_b": 1792,
"successes_shared": 1734,
"success_jaccard": 0.919893899204244,
"answer_agreement": 0.9312072892938497
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 378,
"failures_b": 386,
"failures_shared": 323,
"failure_jaccard": 0.7324263038548753,
"failure_recall_a_in_b": 0.8544973544973545,
"successes_a": 1817,
"successes_b": 1809,
"successes_shared": 1754,
"success_jaccard": 0.936965811965812,
"answer_agreement": 0.9462414578587699
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 378,
"failures_b": 403,
"failures_shared": 325,
"failure_jaccard": 0.7127192982456141,
"failure_recall_a_in_b": 0.8597883597883598,
"successes_a": 1817,
"successes_b": 1792,
"successes_shared": 1739,
"success_jaccard": 0.9299465240641711,
"answer_agreement": 0.9403189066059225
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 386,
"failures_b": 403,
"failures_shared": 336,
"failure_jaccard": 0.7417218543046358,
"failure_recall_a_in_b": 0.8704663212435233,
"successes_a": 1809,
"successes_b": 1792,
"successes_shared": 1742,
"success_jaccard": 0.9370629370629371,
"answer_agreement": 0.9466970387243736
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B-Base",
"task": "youtube_level1",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 821,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 821,
"accuracy": 0.5420219244823387,
"num_failures": 376,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 821,
"accuracy": 0.5456760048721072,
"num_failures": 373,
"flip_rate_vs_reference": 0.10840438489646773,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 821,
"accuracy": 0.5602923264311814,
"num_failures": 361,
"flip_rate_vs_reference": 0.11693057247259439,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 821,
"accuracy": 0.5724725943970768,
"num_failures": 351,
"flip_rate_vs_reference": 0.13276492082825822,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 821,
"accuracy": 0.5663824604141291,
"num_failures": 356,
"flip_rate_vs_reference": 0.14981729598051158,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 821,
"accuracy": 0.5554202192448234,
"num_failures": 365,
"flip_rate_vs_reference": 0.15712545676004872,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.030450669914738104,
"stability": {
"always_correct": 364,
"always_wrong": 286,
"unstable": 171,
"always_correct_rate": 0.44336175395858707,
"always_wrong_rate": 0.34835566382460414,
"unstable_rate": 0.20828258221680876
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 376,
"failures_b": 373,
"failures_shared": 341,
"failure_jaccard": 0.8357843137254902,
"failure_recall_a_in_b": 0.9069148936170213,
"successes_a": 445,
"successes_b": 448,
"successes_shared": 413,
"success_jaccard": 0.8604166666666667,
"answer_agreement": 0.8915956151035322
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 376,
"failures_b": 361,
"failures_shared": 332,
"failure_jaccard": 0.8197530864197531,
"failure_recall_a_in_b": 0.8829787234042553,
"successes_a": 445,
"successes_b": 460,
"successes_shared": 416,
"success_jaccard": 0.8507157464212679,
"answer_agreement": 0.8830694275274056
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 376,
"failures_b": 351,
"failures_shared": 322,
"failure_jaccard": 0.7950617283950617,
"failure_recall_a_in_b": 0.8563829787234043,
"successes_a": 445,
"successes_b": 470,
"successes_shared": 416,
"success_jaccard": 0.8336673346693386,
"answer_agreement": 0.8672350791717418
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 376,
"failures_b": 356,
"failures_shared": 319,
"failure_jaccard": 0.7723970944309927,
"failure_recall_a_in_b": 0.848404255319149,
"successes_a": 445,
"successes_b": 465,
"successes_shared": 408,
"success_jaccard": 0.8127490039840638,
"answer_agreement": 0.8501827040194885
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 376,
"failures_b": 365,
"failures_shared": 320,
"failure_jaccard": 0.7600950118764845,
"failure_recall_a_in_b": 0.851063829787234,
"successes_a": 445,
"successes_b": 456,
"successes_shared": 400,
"success_jaccard": 0.7984031936127745,
"answer_agreement": 0.8428745432399513
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 373,
"failures_b": 361,
"failures_shared": 338,
"failure_jaccard": 0.8535353535353535,
"failure_recall_a_in_b": 0.9061662198391421,
"successes_a": 448,
"successes_b": 460,
"successes_shared": 425,
"success_jaccard": 0.8799171842650103,
"answer_agreement": 0.9037758830694276
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 373,
"failures_b": 351,
"failures_shared": 321,
"failure_jaccard": 0.7965260545905707,
"failure_recall_a_in_b": 0.8605898123324397,
"successes_a": 448,
"successes_b": 470,
"successes_shared": 418,
"success_jaccard": 0.836,
"answer_agreement": 0.8672350791717418
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 373,
"failures_b": 356,
"failures_shared": 318,
"failure_jaccard": 0.7737226277372263,
"failure_recall_a_in_b": 0.8525469168900804,
"successes_a": 448,
"successes_b": 465,
"successes_shared": 410,
"success_jaccard": 0.8151093439363817,
"answer_agreement": 0.8489646772228989
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 373,
"failures_b": 365,
"failures_shared": 319,
"failure_jaccard": 0.7613365155131265,
"failure_recall_a_in_b": 0.8552278820375335,
"successes_a": 448,
"successes_b": 456,
"successes_shared": 402,
"success_jaccard": 0.8007968127490039,
"answer_agreement": 0.8453105968331304
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 361,
"failures_b": 351,
"failures_shared": 318,
"failure_jaccard": 0.8071065989847716,
"failure_recall_a_in_b": 0.8808864265927978,
"successes_a": 460,
"successes_b": 470,
"successes_shared": 427,
"success_jaccard": 0.8489065606361829,
"answer_agreement": 0.876979293544458
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 361,
"failures_b": 356,
"failures_shared": 325,
"failure_jaccard": 0.8290816326530612,
"failure_recall_a_in_b": 0.9002770083102493,
"successes_a": 460,
"successes_b": 465,
"successes_shared": 429,
"success_jaccard": 0.8649193548387096,
"answer_agreement": 0.8855054811205847
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 361,
"failures_b": 365,
"failures_shared": 318,
"failure_jaccard": 0.7794117647058824,
"failure_recall_a_in_b": 0.8808864265927978,
"successes_a": 460,
"successes_b": 456,
"successes_shared": 413,
"success_jaccard": 0.8210735586481114,
"answer_agreement": 0.8587088915956151
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 351,
"failures_b": 356,
"failures_shared": 320,
"failure_jaccard": 0.8268733850129198,
"failure_recall_a_in_b": 0.9116809116809117,
"successes_a": 470,
"successes_b": 465,
"successes_shared": 434,
"success_jaccard": 0.8662674650698603,
"answer_agreement": 0.8842874543239951
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 351,
"failures_b": 365,
"failures_shared": 325,
"failure_jaccard": 0.8312020460358056,
"failure_recall_a_in_b": 0.9259259259259259,
"successes_a": 470,
"successes_b": 456,
"successes_shared": 430,
"success_jaccard": 0.8669354838709677,
"answer_agreement": 0.8964677222898904
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 356,
"failures_b": 365,
"failures_shared": 324,
"failure_jaccard": 0.8161209068010076,
"failure_recall_a_in_b": 0.9101123595505618,
"successes_a": 465,
"successes_b": 456,
"successes_shared": 424,
"success_jaccard": 0.8531187122736419,
"answer_agreement": 0.8891595615103532
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B-Base",
"task": "youtube_level2",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 764,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 764,
"accuracy": 0.4424083769633508,
"num_failures": 426,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 764,
"accuracy": 0.45287958115183247,
"num_failures": 418,
"flip_rate_vs_reference": 0.13089005235602094,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 764,
"accuracy": 0.450261780104712,
"num_failures": 420,
"flip_rate_vs_reference": 0.13743455497382198,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 764,
"accuracy": 0.45418848167539266,
"num_failures": 417,
"flip_rate_vs_reference": 0.14790575916230367,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 764,
"accuracy": 0.450261780104712,
"num_failures": 420,
"flip_rate_vs_reference": 0.17801047120418848,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 764,
"accuracy": 0.4397905759162304,
"num_failures": 428,
"flip_rate_vs_reference": 0.18848167539267016,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.014397905759162277,
"stability": {
"always_correct": 264,
"always_wrong": 350,
"unstable": 150,
"always_correct_rate": 0.34554973821989526,
"always_wrong_rate": 0.4581151832460733,
"unstable_rate": 0.19633507853403143
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 426,
"failures_b": 418,
"failures_shared": 389,
"failure_jaccard": 0.8549450549450549,
"failure_recall_a_in_b": 0.9131455399061033,
"successes_a": 338,
"successes_b": 346,
"successes_shared": 309,
"success_jaccard": 0.824,
"answer_agreement": 0.8691099476439791
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 426,
"failures_b": 420,
"failures_shared": 388,
"failure_jaccard": 0.8471615720524017,
"failure_recall_a_in_b": 0.9107981220657277,
"successes_a": 338,
"successes_b": 344,
"successes_shared": 306,
"success_jaccard": 0.8138297872340425,
"answer_agreement": 0.862565445026178
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 426,
"failures_b": 417,
"failures_shared": 384,
"failure_jaccard": 0.8366013071895425,
"failure_recall_a_in_b": 0.9014084507042254,
"successes_a": 338,
"successes_b": 347,
"successes_shared": 305,
"success_jaccard": 0.8026315789473685,
"answer_agreement": 0.8520942408376964
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 426,
"failures_b": 420,
"failures_shared": 382,
"failure_jaccard": 0.8232758620689655,
"failure_recall_a_in_b": 0.8967136150234741,
"successes_a": 338,
"successes_b": 344,
"successes_shared": 300,
"success_jaccard": 0.7853403141361257,
"answer_agreement": 0.8219895287958116
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 426,
"failures_b": 428,
"failures_shared": 381,
"failure_jaccard": 0.8054968287526427,
"failure_recall_a_in_b": 0.8943661971830986,
"successes_a": 338,
"successes_b": 336,
"successes_shared": 291,
"success_jaccard": 0.7597911227154047,
"answer_agreement": 0.8115183246073299
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 418,
"failures_b": 420,
"failures_shared": 388,
"failure_jaccard": 0.8622222222222222,
"failure_recall_a_in_b": 0.9282296650717703,
"successes_a": 346,
"successes_b": 344,
"successes_shared": 314,
"success_jaccard": 0.8351063829787234,
"answer_agreement": 0.8808900523560209
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 418,
"failures_b": 417,
"failures_shared": 390,
"failure_jaccard": 0.8764044943820225,
"failure_recall_a_in_b": 0.9330143540669856,
"successes_a": 346,
"successes_b": 347,
"successes_shared": 319,
"success_jaccard": 0.8529411764705882,
"answer_agreement": 0.8795811518324608
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 418,
"failures_b": 420,
"failures_shared": 384,
"failure_jaccard": 0.8458149779735683,
"failure_recall_a_in_b": 0.9186602870813397,
"successes_a": 346,
"successes_b": 344,
"successes_shared": 310,
"success_jaccard": 0.8157894736842105,
"answer_agreement": 0.8416230366492147
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 418,
"failures_b": 428,
"failures_shared": 383,
"failure_jaccard": 0.8272138228941684,
"failure_recall_a_in_b": 0.916267942583732,
"successes_a": 346,
"successes_b": 336,
"successes_shared": 301,
"success_jaccard": 0.7900262467191601,
"answer_agreement": 0.8350785340314136
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 420,
"failures_b": 417,
"failures_shared": 397,
"failure_jaccard": 0.9022727272727272,
"failure_recall_a_in_b": 0.9452380952380952,
"successes_a": 344,
"successes_b": 347,
"successes_shared": 324,
"success_jaccard": 0.8828337874659401,
"answer_agreement": 0.9070680628272252
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 420,
"failures_b": 420,
"failures_shared": 391,
"failure_jaccard": 0.8708240534521158,
"failure_recall_a_in_b": 0.930952380952381,
"successes_a": 344,
"successes_b": 344,
"successes_shared": 315,
"success_jaccard": 0.8445040214477212,
"answer_agreement": 0.8638743455497382
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 420,
"failures_b": 428,
"failures_shared": 386,
"failure_jaccard": 0.8354978354978355,
"failure_recall_a_in_b": 0.919047619047619,
"successes_a": 344,
"successes_b": 336,
"successes_shared": 302,
"success_jaccard": 0.798941798941799,
"answer_agreement": 0.8520942408376964
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 417,
"failures_b": 420,
"failures_shared": 393,
"failure_jaccard": 0.8851351351351351,
"failure_recall_a_in_b": 0.9424460431654677,
"successes_a": 347,
"successes_b": 344,
"successes_shared": 320,
"success_jaccard": 0.862533692722372,
"answer_agreement": 0.8808900523560209
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 417,
"failures_b": 428,
"failures_shared": 391,
"failure_jaccard": 0.8612334801762115,
"failure_recall_a_in_b": 0.9376498800959233,
"successes_a": 347,
"successes_b": 336,
"successes_shared": 310,
"success_jaccard": 0.8310991957104558,
"answer_agreement": 0.8730366492146597
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 420,
"failures_b": 428,
"failures_shared": 395,
"failure_jaccard": 0.8719646799116998,
"failure_recall_a_in_b": 0.9404761904761905,
"successes_a": 344,
"successes_b": 336,
"successes_shared": 311,
"success_jaccard": 0.8428184281842819,
"answer_agreement": 0.8848167539267016
}
]
},
{
"model_id": "Qwen/Qwen3.5-4B-Base",
"task": "youtube_level3",
"baselines": [
"0.00",
"0.05",
"0.10",
"0.15",
"0.20",
"0.25"
],
"reference_baseline": "0.00",
"num_common_samples": 836,
"dropped_per_baseline": {
"0.00": 0,
"0.05": 0,
"0.10": 0,
"0.15": 0,
"0.20": 0,
"0.25": 0
},
"per_baseline": [
{
"baseline": "0.00",
"num_samples": 836,
"accuracy": 0.3026315789473684,
"num_failures": 583,
"flip_rate_vs_reference": 0.0,
"unparsed_rate": 0.0
},
{
"baseline": "0.05",
"num_samples": 836,
"accuracy": 0.2715311004784689,
"num_failures": 609,
"flip_rate_vs_reference": 0.24760765550239233,
"unparsed_rate": 0.0
},
{
"baseline": "0.10",
"num_samples": 836,
"accuracy": 0.27751196172248804,
"num_failures": 604,
"flip_rate_vs_reference": 0.2452153110047847,
"unparsed_rate": 0.0
},
{
"baseline": "0.15",
"num_samples": 836,
"accuracy": 0.26674641148325356,
"num_failures": 613,
"flip_rate_vs_reference": 0.26555023923444976,
"unparsed_rate": 0.0
},
{
"baseline": "0.20",
"num_samples": 836,
"accuracy": 0.2727272727272727,
"num_failures": 608,
"flip_rate_vs_reference": 0.27631578947368424,
"unparsed_rate": 0.0
},
{
"baseline": "0.25",
"num_samples": 836,
"accuracy": 0.2822966507177033,
"num_failures": 600,
"flip_rate_vs_reference": 0.284688995215311,
"unparsed_rate": 0.0
}
],
"accuracy_spread": 0.03588516746411485,
"stability": {
"always_correct": 137,
"always_wrong": 482,
"unstable": 217,
"always_correct_rate": 0.1638755980861244,
"always_wrong_rate": 0.5765550239234449,
"unstable_rate": 0.2595693779904306
},
"pairwise": [
{
"baseline_a": "0.00",
"baseline_b": "0.05",
"failures_a": 583,
"failures_b": 609,
"failures_shared": 545,
"failure_jaccard": 0.8423493044822257,
"failure_recall_a_in_b": 0.934819897084048,
"successes_a": 253,
"successes_b": 227,
"successes_shared": 189,
"success_jaccard": 0.6494845360824743,
"answer_agreement": 0.7523923444976076
},
{
"baseline_a": "0.00",
"baseline_b": "0.10",
"failures_a": 583,
"failures_b": 604,
"failures_shared": 541,
"failure_jaccard": 0.8374613003095975,
"failure_recall_a_in_b": 0.9279588336192109,
"successes_a": 253,
"successes_b": 232,
"successes_shared": 190,
"success_jaccard": 0.6440677966101694,
"answer_agreement": 0.7547846889952153
},
{
"baseline_a": "0.00",
"baseline_b": "0.15",
"failures_a": 583,
"failures_b": 613,
"failures_shared": 542,
"failure_jaccard": 0.8287461773700305,
"failure_recall_a_in_b": 0.9296740994854202,
"successes_a": 253,
"successes_b": 223,
"successes_shared": 182,
"success_jaccard": 0.6190476190476191,
"answer_agreement": 0.7344497607655502
},
{
"baseline_a": "0.00",
"baseline_b": "0.20",
"failures_a": 583,
"failures_b": 608,
"failures_shared": 534,
"failure_jaccard": 0.8127853881278538,
"failure_recall_a_in_b": 0.9159519725557461,
"successes_a": 253,
"successes_b": 228,
"successes_shared": 179,
"success_jaccard": 0.5927152317880795,
"answer_agreement": 0.7236842105263158
},
{
"baseline_a": "0.00",
"baseline_b": "0.25",
"failures_a": 583,
"failures_b": 600,
"failures_shared": 528,
"failure_jaccard": 0.8061068702290076,
"failure_recall_a_in_b": 0.9056603773584906,
"successes_a": 253,
"successes_b": 236,
"successes_shared": 181,
"success_jaccard": 0.5876623376623377,
"answer_agreement": 0.715311004784689
},
{
"baseline_a": "0.05",
"baseline_b": "0.10",
"failures_a": 609,
"failures_b": 604,
"failures_shared": 563,
"failure_jaccard": 0.8661538461538462,
"failure_recall_a_in_b": 0.9244663382594417,
"successes_a": 227,
"successes_b": 232,
"successes_shared": 186,
"success_jaccard": 0.6813186813186813,
"answer_agreement": 0.7906698564593302
},
{
"baseline_a": "0.05",
"baseline_b": "0.15",
"failures_a": 609,
"failures_b": 613,
"failures_shared": 567,
"failure_jaccard": 0.8656488549618321,
"failure_recall_a_in_b": 0.9310344827586207,
"successes_a": 227,
"successes_b": 223,
"successes_shared": 181,
"success_jaccard": 0.6728624535315985,
"answer_agreement": 0.7727272727272727
},
{
"baseline_a": "0.05",
"baseline_b": "0.20",
"failures_a": 609,
"failures_b": 608,
"failures_shared": 562,
"failure_jaccard": 0.8580152671755725,
"failure_recall_a_in_b": 0.922824302134647,
"successes_a": 227,
"successes_b": 228,
"successes_shared": 181,
"success_jaccard": 0.6605839416058394,
"answer_agreement": 0.7691387559808612
},
{
"baseline_a": "0.05",
"baseline_b": "0.25",
"failures_a": 609,
"failures_b": 600,
"failures_shared": 550,
"failure_jaccard": 0.834597875569044,
"failure_recall_a_in_b": 0.90311986863711,
"successes_a": 227,
"successes_b": 236,
"successes_shared": 177,
"success_jaccard": 0.6188811188811189,
"answer_agreement": 0.7559808612440191
},
{
"baseline_a": "0.10",
"baseline_b": "0.15",
"failures_a": 604,
"failures_b": 613,
"failures_shared": 569,
"failure_jaccard": 0.8780864197530864,
"failure_recall_a_in_b": 0.9420529801324503,
"successes_a": 232,
"successes_b": 223,
"successes_shared": 188,
"success_jaccard": 0.704119850187266,
"answer_agreement": 0.819377990430622
},
{
"baseline_a": "0.10",
"baseline_b": "0.20",
"failures_a": 604,
"failures_b": 608,
"failures_shared": 566,
"failure_jaccard": 0.8761609907120743,
"failure_recall_a_in_b": 0.9370860927152318,
"successes_a": 232,
"successes_b": 228,
"successes_shared": 190,
"success_jaccard": 0.7037037037037037,
"answer_agreement": 0.8038277511961722
},
{
"baseline_a": "0.10",
"baseline_b": "0.25",
"failures_a": 604,
"failures_b": 600,
"failures_shared": 560,
"failure_jaccard": 0.8695652173913043,
"failure_recall_a_in_b": 0.9271523178807947,
"successes_a": 232,
"successes_b": 236,
"successes_shared": 192,
"success_jaccard": 0.6956521739130435,
"answer_agreement": 0.7954545454545454
},
{
"baseline_a": "0.15",
"baseline_b": "0.20",
"failures_a": 613,
"failures_b": 608,
"failures_shared": 570,
"failure_jaccard": 0.8755760368663594,
"failure_recall_a_in_b": 0.9298531810766721,
"successes_a": 223,
"successes_b": 228,
"successes_shared": 185,
"success_jaccard": 0.6954887218045113,
"answer_agreement": 0.8098086124401914
},
{
"baseline_a": "0.15",
"baseline_b": "0.25",
"failures_a": 613,
"failures_b": 600,
"failures_shared": 563,
"failure_jaccard": 0.8661538461538462,
"failure_recall_a_in_b": 0.9184339314845025,
"successes_a": 223,
"successes_b": 236,
"successes_shared": 186,
"success_jaccard": 0.6813186813186813,
"answer_agreement": 0.7894736842105263
},
{
"baseline_a": "0.20",
"baseline_b": "0.25",
"failures_a": 608,
"failures_b": 600,
"failures_shared": 560,
"failure_jaccard": 0.8641975308641975,
"failure_recall_a_in_b": 0.9210526315789473,
"successes_a": 228,
"successes_b": 236,
"successes_shared": 188,
"success_jaccard": 0.6811594202898551,
"answer_agreement": 0.7978468899521531
}
]
}
]
}