{"total":174,"rows":[{"id":"2509.07506","title":"Astra: A Multi-Agent System for GPU Kernel Performance Optimization","created":"2025-09-09","cited_by_count":51,"area":"agents"},{"id":"2407.12391","title":"LLM Inference Serving: Survey of Recent Advances and Opportunities","created":"2024-07-17","cited_by_count":46,"area":null},{"id":"2511.02230","title":"Continuum: Efficient and Robust Multi-Turn LLM Agent Scheduling with KV Cache Time-to-Live","created":"2025-11-04","cited_by_count":44,"area":"agents"},{"id":"2510.12633","title":"Laminar: A Scalable Asynchronous RL Post-Training Framework","created":"2025-10-14","cited_by_count":37,"area":null},{"id":"2502.20576","title":"OmniRouter: Budget and Performance Controllable Multi-LLM Routing","created":"2025-02-27","cited_by_count":35,"area":null},{"id":"2504.02263","title":"MegaScale-Infer: Serving Mixture-of-Experts at Scale with Disaggregated Expert Parallelism","created":"2025-04-03","cited_by_count":32,"area":"moe"},{"id":"2511.14617","title":"Seer: Online Context Learning for Fast Synchronous LLM Reinforcement Learning","created":"2025-11-18","cited_by_count":25,"area":null},{"id":"2505.04021","title":"Prism: Cost-Efficient Multi-LLM Serving via GPU Memory Ballooning","created":"2025-05-06","cited_by_count":24,"area":"inference"},{"id":"2502.00722","title":"Demystifying Cost-Efficiency in LLM Serving over Heterogeneous GPUs","created":"2025-02-02","cited_by_count":22,"area":"inference"},{"id":"2504.15299","title":"D$^{2}$MoE: Dual Routing and Dynamic Scheduling for Efficient On-Device MoE-based LLM Serving","created":"2025-04-17","cited_by_count":20,"area":"moe"},{"id":"2502.00937","title":"ModServe: Modality- and Stage-Aware Resource Disaggregation for Scalable Multimodal Model Serving","created":"2025-02-02","cited_by_count":18,"area":"mllm"},{"id":"2504.20068","title":"JITServe: SLO-aware LLM Serving with Imprecise Request Information","created":"2025-04-24","cited_by_count":17,"area":"inference"},{"id":"2502.15524","title":"HydraServe: Minimizing Cold Start Latency for Serverless LLM Serving in Public Clouds","created":"2025-02-21","cited_by_count":14,"area":"inference"},{"id":"2505.07460","title":"A Survey on Collaborative Mechanisms Between Large and Small Language Models","created":"2025-05-12","cited_by_count":14,"area":null},{"id":"2409.17264","title":"No Request Left Behind: Tackling Heterogeneity in Long-Context LLM Inference with Medha","created":"2024-09-25","cited_by_count":13,"area":null},{"id":"2505.02922","title":"RetroInfer: A Vector Storage Engine for Scalable Long-Context LLM Inference","created":"2025-05-05","cited_by_count":13,"area":"inference"},{"id":"2507.06608","title":"Nexus:Proactive Intra-GPU Disaggregation of Prefill and Decode in LLM Serving","created":"2025-07-09","cited_by_count":13,"area":"inference"},{"id":"2512.03416","title":"TokenScale: Timely and Accurate Autoscaling for Disaggregated LLM Serving with Token Velocity","created":"2025-12-03","cited_by_count":12,"area":"inference"},{"id":"2412.18169","title":"KunServe: Parameter-centric Memory Management for Efficient Memory Overloading Handling in LLM Serving","created":"2024-12-24","cited_by_count":11,"area":"inference"},{"id":"2507.01438","title":"EdgeLoRA: An Efficient Multi-Tenant LLM Serving System on Edge Devices","created":"2025-07-02","cited_by_count":11,"area":"inference"},{"id":"2510.02758","title":"TokenFlow: Responsive LLM Text Streaming Serving under Request Burst via Preemptive Scheduling","created":"2025-10-03","cited_by_count":10,"area":"inference"},{"id":"2606.16824","title":"CacheWise: Understanding Workloads and Optimizing KVCache Management for Efficiently Serving LLM Coding Agents","created":"2026-06-15","cited_by_count":10,"area":"inference"},{"id":"2411.17741","title":"Chameleon: Adaptive Caching and Scheduling for Many-Adapter LLM Inference Environments","created":"2024-11-24","cited_by_count":9,"area":"inference"},{"id":"2412.04504","title":"Multi-Bin Batching for Increasing LLM Inference Throughput","created":"2024-12-03","cited_by_count":9,"area":null},{"id":"2503.15921","title":"SPIN: Accelerating Large Language Model Inference with Heterogeneous Speculative Models","created":"2025-03-20","cited_by_count":9,"area":"inference"},{"id":"2505.09142","title":"ELIS: Efficient LLM Iterative Scheduling System with Response Length Predictor","created":"2025-05-14","cited_by_count":9,"area":"inference"},{"id":"2510.11938","title":"FlexPipe: Adapting Dynamic LLM Serving Through Inflight Pipeline Refactoring in Fragmented Serverless Clusters","created":"2025-10-13","cited_by_count":9,"area":"inference"},{"id":"2501.06709","title":"Mell: Memory-Efficient Large Language Model Serving via Multi-GPU KV Cache Management","created":"2025-01-12","cited_by_count":8,"area":"inference"},{"id":"2512.13525","title":"Janus: Disaggregating Attention and Experts for Scalable MoE Inference","created":"2025-12-15","cited_by_count":8,"area":"moe"},{"id":"2603.17435","title":"ZipServ: Fast and Memory-Efficient LLM Inference with Hardware-Aware Lossless Compression","created":"2026-03-18","cited_by_count":8,"area":"inference"},{"id":"2507.10069","title":"ElasticMM: Efficient Multimodal LLMs Serving with Elastic Multimodal Parallelism","created":"2025-07-14","cited_by_count":7,"area":"mllm"},{"id":"2605.13734","title":"KVServe: Service-Aware KV Cache Compression for Communication-Efficient Disaggregated LLM Serving","created":"2026-05-13","cited_by_count":7,"area":"inference"},{"id":"2504.07494","title":"Apt-Serve: Adaptive Request Scheduling on Hybrid Cache for Scalable LLM Inference Serving","created":"2025-04-10","cited_by_count":6,"area":"inference"},{"id":"2504.09285","title":"DynaServe: Unified and Elastic Execution for Dynamic Disaggregated LLM Serving","created":"2025-04-12","cited_by_count":6,"area":"inference"},{"id":"2506.02006","title":"MorphServe: Efficient and Workload-Aware LLM Serving via Runtime Quantized Layer Swapping and KV Cache Resizing","created":"2025-05-24","cited_by_count":6,"area":"inference"},{"id":"2508.01002","title":"Optimal Scheduling Algorithms for LLM Inference: Theory and Practice","created":"2025-08-01","cited_by_count":6,"area":"inference"},{"id":"2605.00528","title":"SAGA: Workflow-Atomic Scheduling for AI Agent Inference on GPU Clusters","created":"2026-05-01","cited_by_count":6,"area":"agents"},{"id":"2410.14257","title":"Revisiting Service Level Objectives and System Level Metrics in Large Language Model Serving","created":"2024-10-18","cited_by_count":5,"area":"inference"},{"id":"2505.11916","title":"Arrow: Adaptive Scheduling Mechanisms for Disaggregated LLM Inference Architecture","created":"2025-05-17","cited_by_count":5,"area":"inference"},{"id":"2506.04203","title":"Cascadia: An Efficient Cascade Serving System for Large Language Models","created":"2025-06-04","cited_by_count":5,"area":"inference"},{"id":"2506.15155","title":"eLLM: Elastic Memory Management Framework for Efficient LLM Serving","created":"2025-06-18","cited_by_count":5,"area":"inference"},{"id":"2507.17120","title":"BucketServe: Bucket-Based Dynamic Batching for Smart and Efficient LLM Inference Serving","created":"2025-07-23","cited_by_count":5,"area":"inference"},{"id":"2511.00868","title":"FlexiCache: Leveraging Temporal Stability of Attention Heads for Efficient KV Cache Management","created":"2025-11-02","cited_by_count":5,"area":"inference"},{"id":"2409.15104","title":"PecSched: Preemptive and Efficient Cluster Scheduling for LLM Inference","created":"2024-09-23","cited_by_count":4,"area":null},{"id":"2410.17840","title":"Is the GPU Half-Empty or Half-Full? Practical Scheduling Techniques for LLMs","created":"2024-10-23","cited_by_count":4,"area":null},{"id":"2504.20828","title":"Ascendra: Dynamic Request Prioritization for Efficient LLM Serving","created":"2025-04-29","cited_by_count":4,"area":"inference"},{"id":"2505.17074","title":"Semi-Clairvoyant Scheduling of Speculative Decoding Requests to Minimize LLM Inference Latency","created":"2025-05-20","cited_by_count":4,"area":"inference"},{"id":"2507.17133","title":"BrownoutServe: SLO-Aware Inference Serving under Bursty Workloads for MoE-based LLMs","created":"2025-07-23","cited_by_count":4,"area":"moe"},{"id":"2510.02613","title":"ElasticMoE: An Efficient Auto Scaling Method for Mixture-of-Experts Models","created":"2025-10-02","cited_by_count":4,"area":"moe"},{"id":"2512.23858","title":"Yggdrasil: Bridging Dynamic Speculation and Static Runtime for Latency-Optimal Tree-Based LLM Decoding","created":"2025-12-29","cited_by_count":4,"area":"inference"},{"id":"2602.12029","title":"PrefillShare: A Shared Prefill Module for KV Reuse in Multi-LLM Disaggregated Serving","created":"2026-02-12","cited_by_count":4,"area":"inference"},{"id":"2604.05164","title":"Not All Turns Are Equally Hard: Adaptive Thinking Budgets For Efficient Multi-Turn Reasoning in Agents","created":"2026-04-06","cited_by_count":4,"area":"lr"},{"id":"2501.01792","title":"Efficient LLM Inference with Activation Checkpointing and Hybrid Caching","created":"2025-01-03","cited_by_count":3,"area":"inference"},{"id":"2505.08944","title":"Toward Cost-Efficient Serving of Mixture-of-Experts with Asynchrony","created":"2025-05-13","cited_by_count":3,"area":"moe"},{"id":"2507.21276","title":"LeMix: Unified Scheduling for LLM Training and Inference on Multi-GPU Systems","created":"2025-07-28","cited_by_count":3,"area":null},{"id":"2508.15919","title":"HFX: Joint Design of Algorithms and Systems for Multi-SLO Serving and Fast Scaling","created":"2025-08-21","cited_by_count":3,"area":"inference"},{"id":"2509.00195","title":"FastTTS: Accelerating Test-Time Scaling for Edge LLM Reasoning","created":"2025-08-29","cited_by_count":3,"area":"lr"},{"id":"2510.13223","title":"BanaServe: Unified KV Cache and Dynamic Module Migration for Balancing Disaggregated LLM Serving in AI Infrastructure","created":"2025-10-15","cited_by_count":3,"area":"inference"},{"id":"2601.18999","title":"Randomization Boosts KV Caching, Learning Balances Query Load: A Joint Perspective","created":"2026-01-26","cited_by_count":3,"area":"inference"},{"id":"2601.22438","title":"Towards Resiliency in Large Language Model Serving with KevlarFlow","created":"2026-01-30","cited_by_count":3,"area":"inference"},{"id":"2602.12151","title":"OServe: Accelerating LLM Serving via Spatial-Temporal Workload Orchestration","created":"2026-02-12","cited_by_count":3,"area":"inference"},{"id":"2604.26963","title":"MARS: Efficient, Adaptive Co-Scheduling for Heterogeneous Agentic Systems","created":"2026-04-14","cited_by_count":3,"area":"agents"},{"id":"2605.05467","title":"Nitsum: Serving Tiered LLM Requests with Adaptive Tensor Parallelism","created":"2026-05-06","cited_by_count":3,"area":"inference"},{"id":"2608.01651","title":"Bole: Efficient Tree Speculation for Hybrid-Attention Language Models","created":"2026-08-03","cited_by_count":3,"area":"inference"},{"id":"2502.15763","title":"Hybrid Offline-online Scheduling Method for Large Language Model Inference Optimization","created":"2025-02-14","cited_by_count":2,"area":null},{"id":"2503.13737","title":"AccelGen: Heterogeneous SLO-Guaranteed High-Throughput LLM Inference Serving for Diverse Applications","created":"2025-03-17","cited_by_count":2,"area":null},{"id":"2503.17707","title":"PipeBoost: Resilient Pipelined Architecture for Fast Serverless LLM Scaling","created":"2025-03-22","cited_by_count":2,"area":"inference"},{"id":"2504.09590","title":"Efficient LLM Serving on Hybrid Real-time and Best-effort Requests","created":"2025-04-13","cited_by_count":2,"area":"inference"},{"id":"2505.07680","title":"SpecRouter: Adaptive Routing for Multi-Level Speculative Decoding in Large Language Models","created":"2025-05-12","cited_by_count":2,"area":"inference"},{"id":"2508.01744","title":"AGFT: An Adaptive GPU Frequency Tuner for Real-Time LLM Inference Optimization","created":"2025-08-03","cited_by_count":2,"area":null},{"id":"2508.03611","title":"Astrolabe: Balancing Load in LLM Serving with Randomized Prediction-Guided Scheduling","created":"2025-08-05","cited_by_count":2,"area":"inference"},{"id":"2509.04827","title":"VoltanaLLM: Energy-Efficient and SLO-Aware Disaggregated LLM Serving via Adaptive Frequency Control and State-Space Routing","created":"2025-09-05","cited_by_count":2,"area":"inference"},{"id":"2510.03283","title":"MACE: A Hybrid LLM Serving System with Colocated SLO-aware Continuous Retraining Alignment","created":"2025-09-28","cited_by_count":2,"area":"inference"},{"id":"2511.12979","title":"RAGPulse: An Open-Source RAG Workload Trace to Optimize RAG Serving Systems","created":"2025-11-17","cited_by_count":2,"area":"rag"},{"id":"2512.15784","title":"Beyond Training: Enabling Self-Evolution of Agents with MOBIMEM","created":"2025-12-15","cited_by_count":2,"area":"agents"},{"id":"2512.16134","title":"Staggered Batch Scheduling: Co-optimizing Time-to-First-Token and Throughput for High-Efficiency LLM Inference","created":"2025-12-18","cited_by_count":2,"area":null},{"id":"2512.22492","title":"Role-Based Fault Tolerance System for LLM RL Post-Training","created":"2025-12-27","cited_by_count":2,"area":null},{"id":"2602.07878","title":"Rethinking Latency Denial-of-Service: Attacking the LLM Serving Framework, Not the Model","created":"2026-02-08","cited_by_count":2,"area":"inference"},{"id":"2604.09611","title":"Characterizing Performance-Energy Trade-offs of Large Language Models in Multi-Request Workflows","created":"2026-03-12","cited_by_count":2,"area":"inference"},{"id":"2604.09613","title":"Token-Budget-Aware Pool Routing for Cost-Efficient LLM Inference","created":"2026-03-13","cited_by_count":2,"area":"inference"},{"id":"2606.01839","title":"Observation, Not Prediction: Conversation-Level Disaggregated Scheduling for Agentic Serving","created":"2026-06-01","cited_by_count":2,"area":"agents"},{"id":"2606.09061","title":"Fairness-Aware and Latency-Controllable Scheduling for Chunked-Prefill LLM Serving","created":"2026-06-08","cited_by_count":2,"area":"inference"},{"id":"2606.18741","title":"ReMP: Low-Downtime Runtime Model-Parallelism Reconfiguration for LLM Serving","created":"2026-06-17","cited_by_count":2,"area":"inference"},{"id":"2607.04206","title":"Sangam: Efficiently Serving Diffusion LLMs with the AR Stack","created":"2026-07-05","cited_by_count":2,"area":"diffusion"},{"id":"2607.08565","title":"SMetric: Rethink LLM Scheduling for Serving Agents with Balanced Session-centric Scheduling","created":"2026-07-09","cited_by_count":2,"area":"agents"},{"id":"2607.23815","title":"Kalypso: Relational LLM Serving","created":"2026-07-26","cited_by_count":2,"area":"inference"},{"id":"2501.10187","title":"Good things come in small packages: Should we build AI clusters with Lite-GPUs?","created":"2025-01-17","cited_by_count":1,"area":null},{"id":"2501.13111","title":"iServe: An Intent-based Serving System for LLMs","created":"2025-01-08","cited_by_count":1,"area":"inference"},{"id":"2503.09304","title":"Priority-Aware Preemptive Scheduling for Mixed-Priority Workloads in MoE Inference","created":"2025-03-12","cited_by_count":1,"area":"moe"},{"id":"2506.14851","title":"Efficient Serving of LLM Applications with Probabilistic Demand Modeling","created":"2025-06-17","cited_by_count":1,"area":null},{"id":"2507.18006","title":"Unlock the Potential of Fine-grained LLM Serving via Dynamic Module Scaling","created":"2025-07-24","cited_by_count":1,"area":"inference"},{"id":"2510.19366","title":"MoE-Prism: Disentangling Monolithic Experts for Elastic MoE Services via Model-System Co-Designs","created":"2025-10-22","cited_by_count":1,"area":"moe"},{"id":"2511.06247","title":"Optimizing Long-context LLM Serving via Fine-grained Sequence Parallelism","created":"2025-11-09","cited_by_count":1,"area":"inference"},{"id":"2512.19179","title":"CascadeInfer: Length-Aware Scheduling of LLM Serving with Low Latency and Load Balancing","created":"2025-12-22","cited_by_count":1,"area":"inference"},{"id":"2601.08800","title":"MixServe: An Automatic Distributed Serving System for MoE Models with Hybrid Parallelism Based on Fused Communication Algorithm","created":"2026-01-13","cited_by_count":1,"area":"moe"},{"id":"2601.09258","title":"LatencyPrism: Online Non-intrusive Latency Sculpting for SLO-Guaranteed LLM Inference","created":"2026-01-14","cited_by_count":1,"area":null},{"id":"2602.09725","title":"Efficient Remote KV Cache Reuse with GPU-native Video Codec","created":"2026-02-10","cited_by_count":1,"area":"inference"},{"id":"2602.11530","title":"PASCAL: A Phase-Aware Scheduling Algorithm for Serving Reasoning-based Large Language Models","created":"2026-02-12","cited_by_count":1,"area":"lr"},{"id":"2602.11688","title":"GORGO: Online Tuning for Cross-Region Network-Aware LLM Serving","created":"2026-02-12","cited_by_count":1,"area":"inference"},{"id":"2603.12118","title":"Cornserve: A Distributed Serving System for Any-to-Any Multimodal Models","created":"2026-03-12","cited_by_count":1,"area":null}],"next_cursor":"100"}