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
36 results for “keyword:"early stop"” in ICLR 2027. 7 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]PrimeFocus: Harnessing Intrinsic Regularities for Test-Time Policy Optimization in Vision-Language Models▼4.3k2.00
[PDF]Logarithmic Activation Clocks in Diagonal Linear Networks: When SGD Does and Does Not Follow Lasso▼1351.33
[PDF]One Schedule Does Not Fit All: Adaptive Inference Schedules for Diffusion Language Models▼4.3k2.00
[PDF]When Fixed Thresholds Fail: Trace-Grounded Scheduling for Multi-Tenant LLM Fine-Tuning under SLOs▼4.5k2.25
[PDF]Conditional Diffusion without ERM Oracles: From Finite Neural Training to Sampling Guarantees▼5.8k2.33
[PDF]Supervise What You Certify: Stop-Harm Prediction for Early Stopping in Reasoning Models▼5.8k2.33
[PDF]Implicit Sparse Regularization in Momentum Gradient Descent: A Shifted Early Stopping Window Induced by Momentum and Depth▼4.3k2.00
[PDF]Statistical Benefits of Fine-Tuning from Pretrained Initialization in Diagonal Linear Networks▲3.0k2.33
[PDF]Auditing Early Stopping after Deployment: Sequential Inference with Selective Feedback▲3.0k2.33
[PDF]Stop Early, Spend Less: Equal-Accuracy Frame-Budget Saving via Agreement-Gated Adaptive Sampling in Video Language Models▲11.00
[PDF]When to Stop and Which to Return: Trajectory-Level Minimum Bayes Risk Decoding for Diffusion Language Models▲3.0k2.33
Withdrawn or desk rejected (7)
[PDF]Before the Breakthrough: Bottleneck Crossings as Precursors in Long-Horizon Agent RLWithdrawn2.00
[PDF]Cardano: Training Streaming Video Language Models to Know When They Have Seen EnoughWithdrawn1.67