{ "schema_version": 1, "builder_plan": { "scenario": "A self-hosted model uses more compute or memory than your workload can justify.", "change": "Check for a usable pruned checkpoint and runtime support before attempting model surgery.", "baseline": "The original unpruned model at the same precision, context, and load.", "steps": [ "Locate the paper's implementation and check its license, model support, data needs, and hardware requirements. Stop here if you cannot obtain the required artifacts.", "Save a small set of representative product tasks and expected outcomes. Keep evaluation tasks out of any training or tuning, and choose quality and cost limits before running.", "Run the baseline and the proposed change on identical inputs. Keep other settings fixed and repeat paired runs to expose variation.", "Save per-task outputs, failures, timings, and all usage costs. Review regressions before expanding the test." ], "measure": "Task quality, actual GPU memory, serving throughput, and preparation cost.", "decision": "Keep the change only if it meets your preselected quality and latency limits and improves the metric you care about after setup and operating costs. Otherwise retain the baseline. This is an exploration plan, not a validated recommendation." }, "builder_review": null, "arxiv_id": "2609.27988", "title": "Task-Induced Riemannian Metrics for Vision Transformer Feature Spaces", "summary": "The method replaces Euclidean/cosine-similarity heuristics for ViT token importance (as used in ToMe-style pruning) with a low-rank approximation of the task-induced pullback metric, distilled into a 310K-parameter importance head that scores tokens directly from features. Cost is modest: a matrix-free diagnostic requiring a few Jacobian-vector products, randomized power iteration to fit the metric, and the small head at inference, with no backbone fine-tuning. It can fail when the Jacobian spectrum is too spread for a low-rank fit on a given model-decoder pair, in which case a VAE bottleneck on the decoder input is needed and tractability is not guaranteed.", "rationale": "The gain is pair-dependent and the authors themselves show viability varies across backbones, so the reader should run the low-cost kappa_cap diagnostic on their own model-decoder pair before adopting the pruning head.", "technique": "Task-induced low-rank metric for ViT token pruning", "family": "esparsidade_e_poda", "recommendation": "testar", "published": "2026-09-23", "as_of": "2026-10-18", "editorial_status": "indexed", "source_url": "https://arxiv.org/abs/2609.27988", "independent_implementations": 0, "total_implementations": 0, "stars": 0, "citations": 0, "claims": [ { "claim_id": "claim-01", "statement": "The method replaces Euclidean/cosine-similarity heuristics for ViT token importance (as used in ToMe-style pruning) with a low-rank approximation of the task-induced pullback metric, distilled into a 310K-parameter importance head that scores tokens directly from features. Cost is modest: a matrix-free diagnostic requiring a few Jacobian-vector products, randomized power iteration to fit the metric, and the small head at inference, with no backbone fine-tuning. It can fail when the Jacobian spectrum is too spread for a low-rank fit on a given model-decoder pair, in which case a VAE bottleneck on the decoder input is needed and tractability is not guaranteed.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.27988", "source_page": null, "source_excerpt": "" }, { "claim_id": "claim-02", "statement": "Geometric token pruning reduces the additional depth error of ToMe-based token selection by 25% on DPT depth at prune ratio 0.5, without fine-tuning the ViT; the distilled importance head reaches Spearman rho = 0.998 on DINOv2 CLS.", "basis": "inferred", "result": "", "baseline": "", "conditions": "", "source_url": "https://arxiv.org/abs/2609.27988", "source_page": null, "source_excerpt": "" } ], "exposure_map": [ { "dimension": "quality", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compute", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "latency", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "operations", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "compatibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "security", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "data_and_training", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" }, { "dimension": "reproducibility", "basis": "not_evaluated", "finding": "", "interpretation": "", "limitation": "", "next_check": "", "source_url": "", "source_page": null, "source_excerpt": "" } ], "risks": [], "minimum_test": [], "open_questions": [], "independent_tests": [], "lab_reproductions": [], "report_available": false, "validation_tier": "unknown", "evidence_tier": "unknown", "infrastructure_basis": "unknown", "equations": [], "equations_status": "not_fetched", "core_kind": "", "equations_fetched_at": "", "equations_source_url": "", "equations_source_sha256": "", "exposures_reviewed_at": "", "exposures_source_sha256": "" }