{"total":52,"rows":[{"id":"2410.10781","title":"When Attention Sink Emerges in Language Models: An Empirical View","created":"2024-10-14","cited_by_count":208,"area":"inference"},{"id":"2404.14294","title":"A Survey on Efficient Inference for Large Language Models","created":"2024-04-22","cited_by_count":205,"area":"inference"},{"id":"2402.16363","title":"LLM Inference Unveiled: Survey and Roofline Model Insights","created":"2024-02-26","cited_by_count":152,"area":"inference"},{"id":"2405.14366","title":"MiniCache: KV Cache Compression in Depth Dimension for Large Language Models","created":"2024-05-23","cited_by_count":125,"area":"inference"},{"id":"2312.15234","title":"Towards Efficient Generative Large Language Model Serving: A Survey from Algorithms to Systems","created":"2023-12-23","cited_by_count":102,"area":"inference"},{"id":"2502.08773","title":"Universal Model Routing for Efficient LLM Inference","created":"2025-02-12","cited_by_count":86,"area":"inference"},{"id":"2402.09398","title":"Get More with LESS: Synthesizing Recurrence with KV Cache Compression for Efficient LLM Inference","created":"2024-02-14","cited_by_count":75,"area":"inference"},{"id":"2312.04916","title":"EE-LLM: Large-Scale Training and Inference of Early-Exit Large Language Models with 3D Parallelism","created":"2023-12-08","cited_by_count":69,"area":"inference"},{"id":"2305.17328","title":"Zero-TPrune: Zero-Shot Token Pruning through Leveraging of the Attention Graph in Pre-Trained Transformers","created":"2023-05-27","cited_by_count":62,"area":"inference"},{"id":"2408.12320","title":"TensorOpera Router: A Multi-Model Router for Efficient LLM Inference","created":"2024-08-22","cited_by_count":60,"area":"inference"},{"id":"2408.14690","title":"Training-Free Activation Sparsity in Large Language Models","created":"2024-08-26","cited_by_count":55,"area":"inference"},{"id":"2410.04466","title":"Large Language Model Inference Acceleration: A Comprehensive Hardware Perspective","created":"2024-10-06","cited_by_count":38,"area":"inference"},{"id":"2502.15470","title":"PAPI: Exploiting Dynamic Parallelism in Large Language Model Decoding with a Processing-In-Memory-Enabled Computing System","created":"2025-02-21","cited_by_count":20,"area":"inference"},{"id":"2504.19720","title":"Taming the Titans: A Survey of Efficient LLM Inference Serving","created":"2025-04-28","cited_by_count":16,"area":"inference"},{"id":"2312.05385","title":"Apparate: Rethinking Early Exits to Tame Latency-Throughput Tensions in ML Serving","created":"2023-12-08","cited_by_count":15,"area":"inference"},{"id":"2406.01698","title":"Demystifying AI Platform Design for Distributed Inference of Next-Generation LLM models","created":"2024-06-03","cited_by_count":12,"area":"inference"},{"id":"2509.26182","title":"Parallax: Efficient LLM Inference Service over Decentralized Environment","created":"2025-09-30","cited_by_count":12,"area":"inference"},{"id":"2501.08192","title":"PRESERVE: Prefetching Model Weights and KV-Cache in Distributed LLM Serving","created":"2025-01-14","cited_by_count":9,"area":"inference"},{"id":"2502.16880","title":"CORAL: Learning Consistent Representations across Multi-step Training with Lighter Speculative Drafter","created":"2025-02-24","cited_by_count":9,"area":"inference"},{"id":"2601.06288","title":"AIConfigurator: Lightning-Fast Configuration Optimization for Multi-Framework LLM Serving","created":"2026-01-09","cited_by_count":9,"area":"inference"},{"id":"2410.15252","title":"Lossless KV Cache Compression to 2%","created":"2024-10-20","cited_by_count":7,"area":"inference"},{"id":"2606.06256","title":"RedKnot: Efficient Long-Context LLM Serving with Head-Aware KV Reuse and SegPagedAttention","created":"2026-06-04","cited_by_count":7,"area":"inference"},{"id":"2502.15572","title":"DReSD: Dense Retrieval for Speculative Decoding","created":"2025-02-21","cited_by_count":6,"area":"inference"},{"id":"2507.23777","title":"XSpecMesh: Quality-Preserving Auto-Regressive Mesh Generation Acceleration via Multi-Head Speculative Decoding","created":"2025-07-31","cited_by_count":6,"area":"inference"},{"id":"2408.08470","title":"Context-Aware Assistant Selection for Improved Inference Acceleration with Large Language Models","created":"2024-08-16","cited_by_count":5,"area":"inference"},{"id":"2504.14775","title":"gLLM: Global Balanced Pipeline Parallelism System for Distributed LLM Serving with Token Throttling","created":"2025-04-21","cited_by_count":5,"area":"inference"},{"id":"2403.17726","title":"Tiny Models are the Computational Saver for Large Models","created":"2024-03-26","cited_by_count":4,"area":"inference"},{"id":"2503.04253","title":"ADOR: A Design Exploration Framework for LLM Serving with Enhanced Latency and Throughput","created":"2025-03-06","cited_by_count":4,"area":"inference"},{"id":"2510.14686","title":"xLLM Technical Report","created":"2025-10-16","cited_by_count":3,"area":"inference"},{"id":"2605.09329","title":"Test-Time Speculation","created":"2026-05-10","cited_by_count":3,"area":"inference"},{"id":"2605.04357","title":"Coral: Cost-Efficient Multi-LLM Serving over Heterogeneous Cloud GPUs","created":"2026-05-05","cited_by_count":2,"area":"inference"},{"id":"2502.13251","title":"Neural Attention Search","created":"2025-02-18","cited_by_count":1,"area":"inference"},{"id":"2602.14452","title":"WiSparse: Boosting LLM Inference Efficiency with Weight-Aware Mixed Activation Sparsity","created":"2026-02-16","cited_by_count":1,"area":"inference"},{"id":"2605.01280","title":"Position: LLM Serving Needs Mathematical Optimization and Algorithmic Foundations, Not Just Heuristics","created":"2026-05-02","cited_by_count":1,"area":"inference"},{"id":"2608.12385","title":"Decode-Branch Transformers: Decoupling the Primary Prefill Path from Additional Decode Computation","created":"2026-07-31","cited_by_count":1,"area":"inference"},{"id":"2609.02737","title":"Language Models Can Control Their Own Attention","created":"2026-09-02","cited_by_count":1,"area":"inference"},{"id":"2505.14864","title":"Balanced and Elastic End-to-end Training of Dynamic LLMs","created":"2025-05-20","cited_by_count":0,"area":"inference"},{"id":"2511.21408","title":"Subjective Depth and Timescale Transformers: Learning Where and When to Compute","created":"2025-11-26","cited_by_count":0,"area":"inference"},{"id":"2511.22481","title":"OmniInfer: System-Wide Acceleration Techniques for Optimizing LLM Serving Throughput and Latency","created":"2025-11-27","cited_by_count":0,"area":"inference"},{"id":"2603.21365","title":"TIDE: Token-Informed Depth Execution for Per-Token Early Exit in LLM Inference","created":"2026-03-22","cited_by_count":0,"area":"inference"},{"id":"2605.06113","title":"Tackling the Data-Parallel Load Balancing Bottleneck in LLM Serving: Practical Online Routing at Scale","created":"2026-05-07","cited_by_count":0,"area":"inference"},{"id":"2606.31093","title":"Omni-Flow: A Unified Workflow Orchestration and Distributed KV Cache Sharing Framework for Multimodal Inference","created":"2026-06-30","cited_by_count":0,"area":"inference"},{"id":"2607.02574","title":"From Tensor Buffer to Distributed Memory Hierarchy: A Survey of KV Cache Management for LLM Serving","created":"2026-06-30","cited_by_count":0,"area":"inference"},{"id":"2607.16488","title":"Auto-Scaling Heterogeneous Neural Processing Units for Energy and Cost-Efficient LLM Serving","created":"2026-07-17","cited_by_count":0,"area":"inference"},{"id":"2607.25498","title":"Beyond Prefill-Decode Disaggregation: Dissecting LLM Inference for Heterogeneous Platforms via Dynamic Operator Scheduling","created":"2026-07-28","cited_by_count":0,"area":"inference"},{"id":"2608.25062","title":"FLINT: Efficiently Leveraging High Bandwidth Flash for Capacity-Scalable LLM Inference Acceleration","created":"2026-08-25","cited_by_count":0,"area":"inference"},{"id":"2608.28911","title":"SemKV: Semantic Mixed-Precision KV Cache Quantization Guided by the Quality Cliff for Long-Context LLM Inference","created":"2026-08-28","cited_by_count":0,"area":"inference"},{"id":"2609.12208","title":"Vortex: Bridging Extreme Compression and Efficient LLM Inference","created":"2026-09-10","cited_by_count":0,"area":"inference"},{"id":"2609.12337","title":"Theoretical Guarantees for One-Shot Magnitude Pruning and Compute-Adaptive Early Exit","created":"2026-09-11","cited_by_count":0,"area":"inference"},{"id":"2609.13682","title":"LayerRoute: Adaptive Layer-Skipping with LoRA-Preserved Quality for Efficient LLM Inference","created":"2026-09-12","cited_by_count":0,"area":"inference"},{"id":"2609.32898","title":"When Less Compute Is More: Adaptive Early Exit Improves Pretrained Outlier Detection","created":"2026-09-26","cited_by_count":0,"area":"inference"},{"id":"2609.37887","title":"Behavioral Capacity Certificates for Quantized Language Models","created":"2026-09-29","cited_by_count":0,"area":"inference"}],"next_cursor":null}