{"total":57,"rows":[{"id":"2207.07061","title":"Confident Adaptive Language Modeling","created":"2022-07-14","cited_by_count":310,"area":"inference"},{"id":"2107.00910","title":"Learned Token Pruning for Transformers","created":"2021-07-02","cited_by_count":139,"area":"inference"},{"id":"2312.04985","title":"SparQ Attention: Bandwidth-Efficient LLM Inference","created":"2023-12-08","cited_by_count":119,"area":"inference"},{"id":"2103.04505","title":"Split Computing and Early Exiting for Deep Learning Applications: Survey and Research Challenges","created":"2021-03-08","cited_by_count":115,"area":"inference"},{"id":"2203.07259","title":"The Optimal BERT Surgeon: Scalable and Accurate Second-Order Pruning for Large Language Models","created":"2022-03-14","cited_by_count":115,"area":"inference"},{"id":"2505.23416","title":"KVzip: Query-Agnostic KV Cache Compression with Context Reconstruction","created":"2025-05-29","cited_by_count":95,"area":"inference"},{"id":"2506.06266","title":"Cartridges: Lightweight and general-purpose long context representations via self-study","created":"2025-06-06","cited_by_count":86,"area":"inference"},{"id":"2310.05424","title":"Fast and Robust Early-Exiting Framework for Autoregressive Language Models with Synchronized Parallel Decoding","created":"2023-10-09","cited_by_count":80,"area":"inference"},{"id":"2410.12388","title":"Prompt Compression for Large Language Models: A Survey","created":"2024-10-16","cited_by_count":67,"area":"inference"},{"id":"2408.03094","title":"500xCompressor: Generalized Prompt Compression for Large Language Models","created":"2024-08-06","cited_by_count":47,"area":"inference"},{"id":"2203.13240","title":"Token Dropping for Efficient BERT Pretraining","created":"2022-03-24","cited_by_count":42,"area":"inference"},{"id":"2105.11618","title":"TR-BERT: Dynamic Token Reduction for Accelerating BERT Inference","created":"2021-05-25","cited_by_count":40,"area":"inference"},{"id":"2405.19261","title":"Faster Cascades via Speculative Decoding","created":"2024-05-29","cited_by_count":38,"area":"inference"},{"id":"2405.20915","title":"Fast yet Safe: Early-Exiting with Risk Control","created":"2024-05-31","cited_by_count":30,"area":"inference"},{"id":"2404.03865","title":"FFN-SkipLLM: A Hidden Gem for Autoregressive Decoding with Adaptive Feed Forward Skipping","created":"2024-04-05","cited_by_count":23,"area":"inference"},{"id":"2508.09442","title":"Shadow in the Cache: Unveiling and Mitigating Privacy Risks of KV-cache in LLM Inference","created":"2025-08-13","cited_by_count":21,"area":"inference"},{"id":"2601.17668","title":"Fast KVzip: Efficient and Accurate LLM Inference with Gated KV Eviction","created":"2026-01-25","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":"2305.12228","title":"Dynamic Transformers Provide a False Sense of Efficiency","created":"2023-05-20","cited_by_count":14,"area":"inference"},{"id":"2510.08907","title":"Autoencoding-Free Context Compression for LLMs via Contextual Semantic Anchors","created":"2025-10-10","cited_by_count":14,"area":"inference"},{"id":"2504.08934","title":"Long Context In-Context Compression by Getting to the Gist of Gisting","created":"2025-04-11","cited_by_count":13,"area":"inference"},{"id":"2511.01815","title":"KV Cache Transform Coding for Compact Storage in LLM Inference","created":"2025-11-03","cited_by_count":13,"area":"inference"},{"id":"2305.15273","title":"Revisiting Token Dropping Strategy in Efficient BERT Pretraining","created":"2023-05-24","cited_by_count":11,"area":"inference"},{"id":"2210.13304","title":"ELMER: A Non-Autoregressive Pre-trained Language Model for Efficient and Effective Text Generation","created":"2022-10-24","cited_by_count":10,"area":"inference"},{"id":"2303.09266","title":"SmartBERT: A Promotion of Dynamic Early Exiting Mechanism for Accelerating BERT Inference","created":"2023-03-16","cited_by_count":9,"area":"inference"},{"id":"2406.07368","title":"When Linear Attention Meets Autoregressive Decoding: Towards More Effective and Efficient Linearized Large Language Models","created":"2024-06-11","cited_by_count":9,"area":"inference"},{"id":"2409.14364","title":"Position IDs Matter: An Enhanced Position Layout for Efficient Context Compression in Large Language Models","created":"2024-09-22","cited_by_count":9,"area":"inference"},{"id":"2512.11920","title":"CXL-SpecKV: A Disaggregated FPGA Speculative KV-Cache for Datacenter LLM Serving","created":"2025-12-11","cited_by_count":8,"area":"inference"},{"id":"2311.08623","title":"DEED: Dynamic Early Exit on Decoder for Accelerating Encoder-Decoder Transformer Models","created":"2023-11-15","cited_by_count":7,"area":"inference"},{"id":"2510.20797","title":"No Mean Feat: Simple, Strong Baselines for Context Compression","created":"2025-10-23","cited_by_count":7,"area":"inference"},{"id":"2508.02053","title":"ProCut: LLM Prompt Compression via Attribution Estimation","created":"2025-08-04","cited_by_count":6,"area":"inference"},{"id":"2306.02307","title":"Finding the SWEET Spot: Analysis and Improvement of Adaptive Inference in Low Resource Settings","created":"2023-06-04","cited_by_count":5,"area":"inference"},{"id":"2404.09221","title":"Exploring and Improving Drafts in Blockwise Parallel Decoding","created":"2024-04-14","cited_by_count":5,"area":"inference"},{"id":"2506.08373","title":"Draft-based Approximate Inference for LLMs","created":"2025-06-10","cited_by_count":5,"area":"inference"},{"id":"2509.19228","title":"CompLLM: Compression for Long Context Q&A","created":"2025-09-23","cited_by_count":5,"area":"inference"},{"id":"2604.09048","title":"Watt Counts: Energy-Aware Benchmark for Sustainable LLM Inference on Heterogeneous GPU Architectures","created":"2026-04-10","cited_by_count":5,"area":"inference"},{"id":"2410.18952","title":"Dynamic Vocabulary Pruning in Early-Exit LLMs","created":"2024-10-24","cited_by_count":4,"area":"inference"},{"id":"2502.01025","title":"Knowing When to Stop: Efficient Context Processing via Latent Sufficiency Signals","created":"2025-02-03","cited_by_count":4,"area":"inference"},{"id":"2509.10798","title":"Judge Q: Trainable Queries for Optimized Information Retention in KV Cache Eviction","created":"2025-09-13","cited_by_count":4,"area":"inference"},{"id":"2511.06174","title":"LUT-LLM: Efficient Large Language Model Inference with Memory-based Computations on FPGAs","created":"2025-11-09","cited_by_count":4,"area":"inference"},{"id":"2504.11765","title":"Shared Disk KV Cache Management for Efficient Multi-Instance Inference in RAG-Powered LLMs","created":"2025-04-16","cited_by_count":3,"area":"inference"},{"id":"2603.19733","title":"PoC: Performance-oriented Context Compression for Large Language Models via Performance Prediction","created":"2026-03-20","cited_by_count":3,"area":"inference"},{"id":"2605.09463","title":"Beyond Position Bias: Shifting Context Compression from Position-Driven to Semantic-Driven","created":"2026-05-10","cited_by_count":3,"area":"inference"},{"id":"2307.15517","title":"A Dataflow Compiler for Efficient LLM Inference using Custom Microscaling Formats","created":"2023-07-28","cited_by_count":2,"area":"inference"},{"id":"2504.21228","title":"CachePrune: Teaching LLMs What Not to Follow via KV-Cache Editing","created":"2025-04-29","cited_by_count":2,"area":"inference"},{"id":"2607.19957","title":"HijackKV: New Threat in Position-Independent KV Cache Reuse","created":"2026-07-22","cited_by_count":2,"area":"inference"},{"id":"2602.01778","title":"Data Distribution Matters: A Data-Centric Perspective on Context Compression for Large Language Model","created":"2026-02-02","cited_by_count":1,"area":"inference"},{"id":"2606.17949","title":"RouteBalance: Fused Model Routing and Load Balancing for Heterogeneous LLM Serving","created":"2026-06-16","cited_by_count":1,"area":"inference"},{"id":"2609.25537","title":"Compressing Long Context into Answer-Aligned Memory Embeddings for LLM Inference","created":"2026-09-22","cited_by_count":1,"area":"inference"},{"id":"2609.39329","title":"PatchKV: Weight-Space Compensation of KV Cache","created":"2026-09-30","cited_by_count":1,"area":"inference"},{"id":"2603.25926","title":"Density-aware Soft Context Compression with Semi-Dynamic Compression Ratio","created":"2026-03-26","cited_by_count":0,"area":"inference"},{"id":"2604.16395","title":"Stream2LLM: Overlap Context Streaming and Prefill for Reduced Time-to-First-Token (TTFT)","created":"2026-03-29","cited_by_count":0,"area":"inference"},{"id":"2606.01336","title":"LongAttnComp: Cross-Family Context Compression for Long-Context Reasoning","created":"2026-05-31","cited_by_count":0,"area":"inference"},{"id":"2606.04581","title":"Multi-SPIN: Multi-Access Speculative Inference for Cooperative Token Generation at the Edge","created":"2026-06-03","cited_by_count":0,"area":"inference"},{"id":"2607.14327","title":"PReM: Learning What to Preserve and When to Refresh for Context Compression","created":"2026-07-15","cited_by_count":0,"area":"inference"},{"id":"2609.07966","title":"MetaKV: Adaptive KV Cache Compression for Constrained LLM Inference","created":"2026-09-07","cited_by_count":0,"area":"inference"},{"id":"2609.26100","title":"TSS: Target-Side Sparsification for Speculative Decoding in Domain-Specific Large Language Models","created":"2026-08-15","cited_by_count":0,"area":"inference"}],"next_cursor":null}