| { |
| "schema_version": 2, |
| "title": "Reproduction: Near-Optimal and Efficient First-Order Algorithm for Multi-Task Learning with Shared Linear Representation", |
| "emoji": "🎯", |
| "space_id": "SabaPivot/repro-near-optimal-and-efficient-first-order-algorithm-for-multi-task-learning-with-shared-linea", |
| "paper": { |
| "arxiv_id": "2605.00473" |
| }, |
| "tags": [ |
| "icml2026-repro", |
| "paper-TnquAvyTtL" |
| ], |
| "updated_at": "2026-07-25T04:00:54+00:00", |
| "root": { |
| "slug": "index", |
| "title": "Reproduction: Near-Optimal and Efficient First-Order Algorithm for Multi-Task Learning with Shared Linear Representation", |
| "file": "pages/index.md", |
| "children": [ |
| { |
| "slug": "executive-summary", |
| "title": "Executive summary", |
| "file": "pages/executive-summary/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-1-tpgd-algorithm-1-is-a-two-phase-first-order-gradient-descent-method", |
| "title": "Claim 1: TPGD (Algorithm 1) is a two-phase first-order gradient-descent method", |
| "file": "pages/claim-1-tpgd-algorithm-1-is-a-two-phase-first-order-gradient-descent-method/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-2-phase-i-performs-k1-2-iterations-of-unregularized-gradient-descent", |
| "title": "Claim 2: Phase I performs K1/2 iterations of unregularized gradient descent", |
| "file": "pages/claim-2-phase-i-performs-k1-2-iterations-of-unregularized-gradient-descent/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-3-theorem-5-1-and-corollary-5-3-prove-the-method-attains", |
| "title": "Claim 3: Theorem 5.1 and Corollary 5.3 prove the method attains", |
| "file": "pages/claim-3-theorem-5-1-and-corollary-5-3-prove-the-method-attains/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-4-the-algorithm-achieves-1-iteration-complexity-i-e-convergence-in", |
| "title": "Claim 4: The algorithm achieves Õ(1) iteration complexity, i.e., convergence in", |
| "file": "pages/claim-4-the-algorithm-achieves-1-iteration-complexity-i-e-convergence-in/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-5-the-estimation-error-guarantee-requires-a-per-task-sample-size-of", |
| "title": "Claim 5: The estimation-error guarantee requires a per-task sample size of", |
| "file": "pages/claim-5-the-estimation-error-guarantee-requires-a-per-task-sample-size-of/page.md", |
| "children": [] |
| }, |
| { |
| "slug": "claim-6-theorem-5-4-establishes-excess-risk-bounds-for-transferring-the-learned", |
| "title": "Claim 6: Theorem 5.4 establishes excess-risk bounds for transferring the learned", |
| "file": "pages/claim-6-theorem-5-4-establishes-excess-risk-bounds-for-transferring-the-learned/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": 8404, |
| "trace_view_tokens": 1850621, |
| "workspace_view_tokens": 309, |
| "revision": "978198954c593c9f2233", |
| "traces_ref": { |
| "repo_id": "SabaPivot/icml26-tnquavyttl-traces", |
| "repo_type": "dataset", |
| "repo_url": "https://huggingface.co/datasets/SabaPivot/icml26-tnquavyttl-traces", |
| "private": true |
| }, |
| "trace_dataset": "https://huggingface.co/datasets/SabaPivot/icml26-tnquavyttl-traces", |
| "workspace_ref": { |
| "repo_id": "SabaPivot/icml26-tnquavyttl-artifacts", |
| "repo_type": "bucket", |
| "repo_url": "https://huggingface.co/buckets/SabaPivot/icml26-tnquavyttl-artifacts", |
| "private": true |
| }, |
| "workspace_bucket": "https://huggingface.co/buckets/SabaPivot/icml26-tnquavyttl-artifacts" |
| } |
