NeurIPS 2026 reviews are up: reviews, scores, rebuttals and meta-reviews for 7,877 papers. See the highest rated
acceptodds: the papers of ICLR 2027
39 results for “keyword:"certificates"” in ICLR 2027. 7 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Where Does Task Identity Live? FLAT: A Provenance-Complete Fixed-Log Audit for Offline Task Identification▼4.3k2.00
[PDF]Which Predictions Suffice for a World-Model Planner's Decision? Certifying Reward and Value Evidence by Re-Planning▲5.4k3.00
[PDF]OWNERSHIP IS THE MISSING DECLARATION: DECIDING RELATIONAL TARGET LEAKAGE FROM THE SCHEMA▼4.3k2.00
[PDF]Your FID Cannot See the Tail: Auditing and Repairing the Rare-Event Calibration of Large Generative Models▲4002.67
[PDF]PACT: Provenance-Aware Contract Testing — What Do LLM-Agent Task-Contract Checkers Actually Certify?▼1351.33
[PDF]Counterfactual Fragility Certificates for Tabular Classifications: Guarantees and Limits▼1.2k1.67
[PDF]From Continuous Policy Optimization to Sparse Deployment: Bellman-Certified Rounding for MDPs▼1.2k3.00
[PDF]Directional Linear Separability of Neural Representations: Geometry and Transformations▼1351.33
Withdrawn or desk rejected (7)
[PDF]Beyond Training Loss: Full-Space Certificates for Neural JKO and Wasserstein EvolutionDesk rejected
[PDF]WHEN BATCH SCHEDULES CANNOT BE CERTIFIED: CAPACITY OF GLOBALLY VALID MONOMIAL CERTIFICATES ON INTEGER LOG-HULLSWithdrawn1.33
[PDF]Certified Feasibility Reporting Under Partial Observability: Distinguishing Unreachable from Unresolved in Tool-Using AgentsDesk rejected