SabaPivot's picture
Publish canonical reproduction with fresh CPU audit
b154008 verified
Raw
History Blame Contribute Delete
3.61 kB
{
"schema_version": 2,
"title": "CodeTaste: Can LLMs Generate Human-Level Code Refactorings?",
"emoji": "🎯",
"space_id": "SabaPivot/repro-codetaste-can-llms-generate-human-level-code-refactorings",
"paper": {
"arxiv_id": "2603.04177"
},
"tags": [
"icml2026-repro",
"paper-nj1oXU2tln"
],
"updated_at": "2026-08-01T09:32:18.729961+00:00",
"root": {
"slug": "index",
"title": "CodeTaste: Can LLMs Generate Human-Level Code Refactorings?",
"file": "pages/index.md",
"children": [
{
"slug": "executive-summary",
"title": "Executive summary",
"file": "pages/executive-summary/page.md",
"children": []
},
{
"slug": "claim-1-benchmark-composition-and-task-scale",
"title": "Claim 1: Benchmark composition and task scale",
"file": "pages/claim-1-benchmark-composition-and-task-scale/page.md",
"children": []
},
{
"slug": "claim-2-instructed-track-leaderboard",
"title": "Claim 2: Instructed Track leaderboard",
"file": "pages/claim-2-instructed-track-leaderboard/page.md",
"children": []
},
{
"slug": "claim-3-open-track-alignment-collapse",
"title": "Claim 3: Open Track alignment collapse",
"file": "pages/claim-3-open-track-alignment-collapse/page.md",
"children": []
},
{
"slug": "claim-4-plan-mode-and-oracle-selection-gains",
"title": "Claim 4: Plan mode and oracle selection gains",
"file": "pages/claim-4-plan-mode-and-oracle-selection-gains/page.md",
"children": []
},
{
"slug": "claim-5-static-rule-evaluation-scale",
"title": "Claim 5: Static rule evaluation scale",
"file": "pages/claim-5-static-rule-evaluation-scale/page.md",
"children": []
},
{
"slug": "claim-6-agent-inference-cost",
"title": "Claim 6: Agent inference cost",
"file": "pages/claim-6-agent-inference-cost/page.md",
"children": []
},
{
"slug": "claim-99-fresh-independent-cpu-audit",
"title": "Fresh independent CPU audit",
"file": "pages/claim-99-fresh-independent-cpu-audit/page.md",
"children": []
},
{
"slug": "conclusion",
"title": "Conclusion",
"file": "pages/conclusion/page.md",
"children": []
}
]
},
"traces": [],
"workspace": {
"file": "workspace.json",
"file_count": 0,
"total_size": 0,
"bucket_id": null
},
"agent_view_tokens": 4836,
"trace_view_tokens": 43554,
"workspace_view_tokens": 5685,
"revision": "4f733a8f6715cfbf0e81",
"traces_ref": {
"repo_id": "Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-traces",
"repo_type": "dataset",
"repo_url": "https://huggingface.co/datasets/Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-traces",
"private": true
},
"trace_dataset": "https://huggingface.co/datasets/Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-traces",
"workspace_ref": {
"repo_id": "Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-artifacts",
"repo_type": "bucket",
"repo_url": "https://huggingface.co/buckets/Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-artifacts",
"private": true
},
"workspace_bucket": "https://huggingface.co/buckets/Srishti280992/repro-codetaste-can-llms-generate-human-level-code-refactorings-artifacts"
}