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
814 results for “keyword:"test time"” in ICLR 2027. 195 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]Regime-Dependent Inference-Time Compute Allocation for Agentic Systems: When Deeper Reasoning Wins▲3.0k2.33
[PDF]Compile the Catalog, Not the Task: ToolGraph as a Tool-Use Agent Harness that Updates, Scales, and Transfers▲6.1k2.00
[PDF]PolicyTTT: The Policy as the Product of Test-Time Training for Clinical Rubric Generation▼4.3k2.00
[PDF]Coverage-Constrained Neural Exponential Embedding: Exact Test-Time Projection for Frozen Generators▼1361.33
[PDF]From Teacher Imitation to Failure Diagnosis: Causality-Inspired Test-Time Evolution of Small Language Models with LLM Feedback▼1.2k1.67
[PDF]Don't Fix What Isn't Broken: Calibrated Repair of Frozen Foundation Models under Product Shift▼4.5k2.67
[PDF]ReDraft: Your Verifier Is Secretly a Depth Supervisor for Recurrent Speculative Decoding▲5103.00
[PDF]Sample the Order, Not the Token: Test-Time Scaling for Masked Diffusion Language Models▲3952.67
Withdrawn or desk rejected (195)
[PDF]Beyond Online Accuracy: What Persists After Continual Test-Time Adaptation of Multimodal LLMs on Multiple-Choice VQA?Withdrawn1.67
[PDF]LongVU-TTT: Causal Test-Time Training for Visual Resampling in Long Video UnderstandingWithdrawn1.75
[PDF]BasePrompt: Context as Prompt for Causal Genome Language Models in Zero-Shot RNA Fitness PredictionWithdrawn1.67
[PDF]When Should the Count Change? Learning State Maintenance for Causal Video CountingDesk rejected2.00