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
34 results for “keyword:"latency"” in ICLR 2027. 7 withdrawn or desk rejected listed apart. clearsort:relevanceratingnewestmost read
[PDF]When to Wait Before Fan-Out: Warm-Up, Verification, and Residency in Cache-Aware LLM Evaluation▲9.2k2.67
[PDF]Safe to Omit, Worth Omitting? Keeping the Answer and Cutting Latency When an LLM Agent Skips Another's Message▲13k2.33
[PDF]World Models in Robotics and Autonomous Driving: A Survey of the Latency-Fidelity Trade-off and the Case for Unified Scene-Motion Dynamics▲11.00
[PDF]SAT-LLM: Conversion-Aware Fine-Tuning of Large Language Models for Low-Latency Spiking Deployment▼1.2k1.67
[PDF]Muscles Know First: Decoding Hand Motion from Wrist EMG for Latency-Compensated Vision-Based Robot Teleoperation▲4002.67
[PDF]EXACT WARM-STATE REUSE FOR HYBRID DIFFUSION LANGUAGE MODELS: CONTRACTS, SPEEDUPS, AND FAILURE BOUNDARIES▲3.0k2.33
[PDF]Every Second Counts: Quality-Cost-Aware LLM Routing for Low-Latency Concurrent Multi-LLM Serving▼1.2k1.67
[PDF]Does the World Wait While Agents Think? A Temporal Fidelity Audit of Interactive-Agent Benchmarks▲9.2k2.67
[PDF]Prune Less, Pay More: Imperceptible Budget Inflation Attacks on Token Pruning in Vision-Language Models▼1.2k1.67