Skip to content

feat: add results for gemini-3.1-pro, gpt-5.5, opus-4.6 #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
Khurdhula-Harshavardhan merged 1 commit into from
Apr 30, 2026
Merged

feat: add results for gemini-3.1-pro, gpt-5.5, opus-4.6 #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
3 changes: 3 additions & 0 deletions data/audio_responses/response_claude-opus-4-6_audio.jsonl
View file
Open in desktop
Git LFS file not shown
3 changes: 3 additions & 0 deletions data/audio_responses/response_gemini-3.1-pro-preview_audio.jsonl
View file
Open in desktop
Git LFS file not shown
3 changes: 3 additions & 0 deletions data/audio_responses/response_gpt-5.5_audio.jsonl
View file
Open in desktop
Git LFS file not shown
3 changes: 3 additions & 0 deletions data/evaluation/audio/claude-opus-4-6/eval_records.jsonl
View file
Open in desktop
Git LFS file not shown
264 changes: 264 additions & 0 deletions data/evaluation/audio/claude-opus-4-6/eval_summary.json
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -0,0 +1,264 @@
{
"response_file": "data/audio_responses/response_claude-opus-4-6_audio.jsonl",
"num_records": 115,
"model_ids": [
"claude-opus-4-6"
],
"data_quality": {
"json_parse_fail_count": 3,
"json_non_structured_root_count": 3,
"invalid_schema_input_count": 0,
"unknown_difficulty_count": 0,
"malformed_jsonl_line_count": 0
},
"summary": {
"overall": {
"n": 115,
"metrics": {
"json_parse_success": {
"mean": 0.9739130434782609,
"ci95_low": 0.9391304347826087,
"ci95_high": 1.0,
"metric_name": "JSON Parse Success"
},
"json_root_structured": {
"mean": 0.9739130434782609,
"ci95_low": 0.9391304347826087,
"ci95_high": 1.0,
"metric_name": "Structured JSON Root"
},
"schema_valid_input": {
"mean": 1.0,
"ci95_low": 1.0,
"ci95_high": 1.0,
"metric_name": "Schema Valid Input"
},
"schema_compliance": {
"mean": 0.9130434782608695,
"ci95_low": 0.8608695652173913,
"ci95_high": 0.9652173913043478,
"metric_name": "JSON Pass Rate"
},
"leaf_value_em": {
"mean": 0.1919377216490032,
"ci95_low": 0.159491619514694,
"ci95_high": 0.22479062603646935,
"metric_name": "Truth Score"
},
"value_token_f1": {
"mean": 0.4103479834753111,
"ci95_low": 0.37027785814336484,
"ci95_high": 0.4546600357579809,
"metric_name": "Faithfulness Score"
},
"hier_path_recall": {
"mean": 0.8811306621813169,
"ci95_low": 0.8269572435028659,
"ci95_high": 0.928100609363258,
"metric_name": "Path Recall"
},
"path_set_f1": {
"mean": 0.8237838889610478,
"ci95_low": 0.7686515536600586,
"ci95_high": 0.8713310465168905,
"metric_name": "Structure Coverage"
},
"type_precision": {
"mean": 0.9130434782608695,
"ci95_low": 0.8608695652173913,
"ci95_high": 0.9565217391304348,
"metric_name": "Type Safety"
},
"strict_json_em": {
"mean": 0.0,
"ci95_low": 0.0,
"ci95_high": 0.0,
"metric_name": "Perfect Response Rate"
}
},
"category_scores": {
"Long Context Extraction": {
"mean": 0.49447212243521044,
"ci95_low": 0.4546205471493368,
"ci95_high": 0.527625826217163,
"category_name": "Long Context Extraction",
"components": [
"leaf_value_em",
"value_token_f1",
"hier_path_recall"
]
},
"Complex Schema Handling": {
"mean": 0.8832902818275957,
"ci95_low": 0.8302234325250624,
"ci95_high": 0.9278670052940108,
"category_name": "Complex Schema Handling",
"components": [
"schema_compliance",
"path_set_f1",
"type_precision"
]
},
"Multi-Context Linking": {
"mean": 0.3011428525621571,
"ci95_low": 0.2644303737263097,
"ci95_high": 0.3359529331203245,
"category_name": "Multi-Context Linking",
"components": [
"leaf_value_em",
"value_token_f1"
]
},
"Output Contract Reliability": {
"mean": 0.9333333333333332,
"ci95_low": 0.889855072463768,
"ci95_high": 0.9710144927536233,
"category_name": "Output Contract Reliability",
"components": [
"json_parse_success",
"schema_compliance",
"type_precision"
]
},
"Strict Precision": {
"mean": 0.0,
"ci95_low": 0.0,
"ci95_high": 0.0,
"category_name": "Strict Precision",
"components": [
"strict_json_em"
]
}
}
},
"overall_weighted": {
"n": 115,
"metrics": {
"json_parse_success": {
"mean": 0.9737609329446064,
"ci95_low": 0.938953488372093,
"ci95_high": 1.0,
"metric_name": "JSON Parse Success"
},
"json_root_structured": {
"mean": 0.9737609329446064,
"ci95_low": 0.9387755102040817,
"ci95_high": 1.0,
"metric_name": "Structured JSON Root"
},
"schema_valid_input": {
"mean": 1.0,
"ci95_low": 1.0,
"ci95_high": 1.0,
"metric_name": "Schema Valid Input"
},
"schema_compliance": {
"mean": 0.9125364431486881,
"ci95_low": 0.8604651162790697,
"ci95_high": 0.9565217391304348,
"metric_name": "JSON Pass Rate"
},
"leaf_value_em": {
"mean": 0.1919967277121567,
"ci95_low": 0.16396622605746825,
"ci95_high": 0.2230335604540147,
"metric_name": "Truth Score"
},
"value_token_f1": {
"mean": 0.41026156212912757,
"ci95_low": 0.3696409331781785,
"ci95_high": 0.452456382456549,
"metric_name": "Faithfulness Score"
},
"hier_path_recall": {
"mean": 0.8804375465089047,
"ci95_low": 0.8277094612672905,
"ci95_high": 0.9302326868199687,
"metric_name": "Path Recall"
},
"path_set_f1": {
"mean": 0.8235063792080847,
"ci95_low": 0.7758698095456549,
"ci95_high": 0.8660132818524661,
"metric_name": "Structure Coverage"
},
"type_precision": {
"mean": 0.9125364431486881,
"ci95_low": 0.8596491228070176,
"ci95_high": 0.9565217391304348,
"metric_name": "Type Safety"
},
"strict_json_em": {
"mean": 0.0,
"ci95_low": 0.0,
"ci95_high": 0.0,
"metric_name": "Perfect Response Rate"
}
},
"category_scores": {
"Long Context Extraction": {
"mean": 0.494231945450063,
"ci95_low": 0.4568342563057294,
"ci95_high": 0.5299255463448895,
"category_name": "Long Context Extraction",
"components": [
"leaf_value_em",
"value_token_f1",
"hier_path_recall"
]
},
"Complex Schema Handling": {
"mean": 0.882859755168487,
"ci95_low": 0.8286985043230447,
"ci95_high": 0.9312911775642015,
"category_name": "Complex Schema Handling",
"components": [
"schema_compliance",
"path_set_f1",
"type_precision"
]
},
"Multi-Context Linking": {
"mean": 0.30112914492064213,
"ci95_low": 0.26573178083632193,
"ci95_high": 0.3343219021909188,
"category_name": "Multi-Context Linking",
"components": [
"leaf_value_em",
"value_token_f1"
]
},
"Output Contract Reliability": {
"mean": 0.9329446064139941,
"ci95_low": 0.8892128279883382,
"ci95_high": 0.9706744868035191,
"category_name": "Output Contract Reliability",
"components": [
"json_parse_success",
"schema_compliance",
"type_precision"
]
},
"Strict Precision": {
"mean": 0.0,
"ci95_low": 0.0,
"ci95_high": 0.0,
"category_name": "Strict Precision",
"components": [
"strict_json_em"
]
}
},
"weighting": "schema_complexity",
"weight_field_priority": [
"schema_complexity",
"difficulty"
],
"difficulty_weights": {
"easy": 1.0,
"medium": 2.0,
"hard": 3.0
}
}
}
}
3 changes: 3 additions & 0 deletions data/evaluation/audio/gemini-3.1-pro-preview/eval_records.jsonl
View file
Open in desktop
Git LFS file not shown
Loading
Loading