{"total":332,"rows":[{"id":"2210.03350","title":"Measuring and Narrowing the Compositionality Gap in Language Models","created":"2022-10-07","cited_by_count":1075,"area":"lr"},{"id":"2102.04906","title":"Dynamic Neural Networks: A Survey","created":"2021-02-09","cited_by_count":408,"area":null},{"id":"2202.07800","title":"Not All Patches are What You Need: Expediting Vision Transformers via Token Reorganizations","created":"2022-02-16","cited_by_count":363,"area":null},{"id":"2112.07658","title":"AdaViT: Adaptive Tokens for Efficient Vision Transformer","created":"2021-12-14","cited_by_count":362,"area":null},{"id":"2207.07061","title":"Confident Adaptive Language Modeling","created":"2022-07-14","cited_by_count":310,"area":"inference"},{"id":"2404.02258","title":"Mixture-of-Depths: Dynamically allocating compute in transformer-based language models","created":"2024-04-02","cited_by_count":258,"area":null},{"id":"2111.15668","title":"AdaViT: Adaptive Vision Transformers for Efficient Image Recognition","created":"2021-11-30","cited_by_count":228,"area":null},{"id":"2105.15075","title":"Not All Images are Worth 16x16 Words: Dynamic Transformers for Efficient Image Recognition","created":"2021-05-31","cited_by_count":144,"area":null},{"id":"2106.12620","title":"IA-RED$^2$: Interpretability-Aware Redundancy Reduction for Vision Transformers","created":"2021-06-23","cited_by_count":127,"area":null},{"id":"2106.03760","title":"DSelect-k: Differentiable Selection in the Mixture of Experts with Applications to Multi-Task Learning","created":"2021-06-07","cited_by_count":115,"area":"moe"},{"id":"2102.04170","title":"Learning Task-Oriented Communication for Edge Inference: An Information Bottleneck Approach","created":"2021-02-08","cited_by_count":110,"area":null},{"id":"2403.02181","title":"Not All Layers of LLMs Are Necessary During Inference","created":"2024-03-04","cited_by_count":96,"area":null},{"id":"2104.14726","title":"MOOD: Multi-level Out-of-distribution Detection","created":"2021-04-30","cited_by_count":94,"area":null},{"id":"2104.06403","title":"Lite-HRNet: A Lightweight High-Resolution Network","created":"2021-04-13","cited_by_count":91,"area":null},{"id":"2103.15538","title":"SUTD-TrafficQA: A Question Answering Benchmark and an Efficient Network for Video Reasoning over Traffic Events","created":"2021-03-29","cited_by_count":84,"area":null},{"id":"2103.13258","title":"Dynamic Slimmable Network","created":"2021-03-24","cited_by_count":83,"area":null},{"id":"2108.08617","title":"Spatially-Adaptive Image Restoration using Distortion-Guided Networks","created":"2021-08-19","cited_by_count":77,"area":null},{"id":"2106.05022","title":"Adaptive Inference through Early-Exit Networks: Design, Challenges and Directions","created":"2021-06-09","cited_by_count":75,"area":"inference"},{"id":"2104.08803","title":"Consistent Accelerated Inference via Confident Adaptive Transformers","created":"2021-04-18","cited_by_count":74,"area":null},{"id":"2410.03456","title":"Dynamic Diffusion Transformer","created":"2024-10-04","cited_by_count":62,"area":"diffusion"},{"id":"2104.13400","title":"FrameExit: Conditional Early Exiting for Efficient Video Recognition","created":"2021-04-27","cited_by_count":59,"area":null},{"id":"2204.00102","title":"Dynamic Multimodal Fusion","created":"2022-03-31","cited_by_count":55,"area":null},{"id":"2303.09875","title":"A Dynamic Multi-Scale Voxel Flow Network for Video Prediction","created":"2023-03-17","cited_by_count":53,"area":null},{"id":"2111.05193","title":"A Survey on Green Deep Learning","created":"2021-11-08","cited_by_count":52,"area":null},{"id":"2303.07820","title":"Adaptive Rotated Convolution for Rotated Object Detection","created":"2023-03-14","cited_by_count":51,"area":null},{"id":"2203.15859","title":"NICGSlowDown: Evaluating the Efficiency Robustness of Neural Image Caption Generation Models","created":"2022-03-29","cited_by_count":50,"area":null},{"id":"2102.05775","title":"AdaFuse: Adaptive Temporal Fusion Network for Efficient Action Recognition","created":"2021-02-10","cited_by_count":48,"area":null},{"id":"2101.00590","title":"RegNet: Self-Regulated Network for Image Classification","created":"2021-01-03","cited_by_count":44,"area":null},{"id":"2209.08310","title":"Learning to Weight Samples for Dynamic Early-exiting Networks","created":"2022-09-17","cited_by_count":44,"area":"inference"},{"id":"2412.03248","title":"AIM: Adaptive Inference of Multi-Modal LLMs via Token Merging and Pruning","created":"2024-12-04","cited_by_count":43,"area":"mllm"},{"id":"2201.03014","title":"Glance and Focus Networks for Dynamic Visual Recognition","created":"2022-01-09","cited_by_count":41,"area":null},{"id":"2304.01230","title":"SEENN: Towards Temporal Spiking Early-Exit Neural Networks","created":"2023-04-02","cited_by_count":41,"area":null},{"id":"2106.05409","title":"Zero Time Waste: Recycling Predictions in Early Exit Neural Networks","created":"2021-06-09","cited_by_count":39,"area":"inference"},{"id":"2111.06061","title":"Edge-Cloud Polarization and Collaboration: A Comprehensive Survey for AI","created":"2021-11-11","cited_by_count":39,"area":null},{"id":"2211.09961","title":"Path Independent Equilibrium Models Can Better Exploit Test-Time Computation","created":"2022-11-18","cited_by_count":38,"area":null},{"id":"2103.03467","title":"Teachers Do More Than Teach: Compressing Image-to-Image Models","created":"2021-03-05","cited_by_count":37,"area":null},{"id":"2105.05165","title":"AdaMML: Adaptive Multi-Modal Learning for Efficient Video Recognition","created":"2021-05-11","cited_by_count":36,"area":null},{"id":"2108.10394","title":"Dynamic Network Quantization for Efficient Video Inference","created":"2021-08-23","cited_by_count":36,"area":null},{"id":"2110.06178","title":"TAda! Temporally-Adaptive Convolutions for Video Understanding","created":"2021-10-12","cited_by_count":36,"area":null},{"id":"2203.09747","title":"Efficient Split-Mix Federated Learning for On-Demand and In-Situ Customization","created":"2022-03-18","cited_by_count":35,"area":null},{"id":"2404.14408","title":"SpaceByte: Towards Deleting Tokenization from Large Language Modeling","created":"2024-04-22","cited_by_count":34,"area":null},{"id":"2306.11248","title":"Dynamic Perceiver for Efficient Visual Recognition","created":"2023-06-20","cited_by_count":33,"area":null},{"id":"2204.11786","title":"Enable Deep Learning on Mobile Devices: Methods, Systems, and Applications","created":"2022-04-25","cited_by_count":32,"area":null},{"id":"2512.09106","title":"Learning Unmasking Policies for Diffusion Language Models","created":"2025-12-09","cited_by_count":30,"area":"diffusion"},{"id":"2110.08678","title":"Improving Transformers with Probabilistic Attention Keys","created":"2021-10-16","cited_by_count":28,"area":null},{"id":"2105.09121","title":"Single-Layer Vision Transformers for More Accurate Early Exits with Less Overhead","created":"2021-05-19","cited_by_count":26,"area":"inference"},{"id":"2108.05894","title":"MicroNet: Improving Image Recognition with Extremely Low FLOPs","created":"2021-08-12","cited_by_count":26,"area":null},{"id":"2110.04456","title":"Deep Joint Source-Channel Coding for Wireless Image Transmission with Adaptive Rate Control","created":"2021-10-09","cited_by_count":26,"area":null},{"id":"2112.00054","title":"Task2Sim : Towards Effective Pre-training and Transfer from Synthetic Data","created":"2021-11-30","cited_by_count":26,"area":"synth"},{"id":"2308.15949","title":"Latency-aware Unified Dynamic Networks for Efficient Image Recognition","created":"2023-08-30","cited_by_count":26,"area":null},{"id":"2207.10909","title":"DBQ-SSD: Dynamic Ball Query for Efficient 3D Object Detection","created":"2022-07-22","cited_by_count":25,"area":null},{"id":"2210.06223","title":"Latency-aware Spatial-wise Dynamic Networks","created":"2022-10-12","cited_by_count":25,"area":null},{"id":"2312.00333","title":"Green Edge AI: A Contemporary Survey","created":"2023-12-01","cited_by_count":25,"area":null},{"id":"2206.04685","title":"Predictive Exit: Prediction of Fine-Grained Early Exits for Computation- and Energy-Efficient Inference","created":"2022-06-09","cited_by_count":24,"area":"inference"},{"id":"2207.01580","title":"Dynamic Spatial Sparsification for Efficient Vision Transformers and Convolutional Neural Networks","created":"2022-07-04","cited_by_count":24,"area":null},{"id":"2301.03831","title":"Dynamic Grained Encoder for Vision Transformers","created":"2023-01-10","cited_by_count":24,"area":null},{"id":"2402.18447","title":"Prompt-Driven Dynamic Object-Centric Learning for Single Domain Generalization","created":"2024-02-28","cited_by_count":24,"area":null},{"id":"2505.20353","title":"FastCache: Fast Caching for Diffusion Transformer Through Learnable Linear Approximation","created":"2025-05-26","cited_by_count":24,"area":"diffusion"},{"id":"2108.04890","title":"On the Effect of Pruning on Adversarial Robustness","created":"2021-08-10","cited_by_count":23,"area":null},{"id":"2412.16822","title":"Layer- and Timestep-Adaptive Differentiable Token Compression Ratios for Efficient Diffusion Transformers","created":"2024-12-22","cited_by_count":22,"area":"diffusion"},{"id":"2203.14949","title":"Controllable Dynamic Multi-Task Architectures","created":"2022-03-28","cited_by_count":21,"area":null},{"id":"2102.11646","title":"HardCoRe-NAS: Hard Constrained diffeRentiable Neural Architecture Search","created":"2021-02-23","cited_by_count":20,"area":null},{"id":"2105.08714","title":"Fighting Gradients with Gradients: Dynamic Defenses against Adversarial Attacks","created":"2021-05-18","cited_by_count":20,"area":null},{"id":"2205.03055","title":"Continual Object Detection via Prototypical Task Correlation Guided Gating Mechanism","created":"2022-05-06","cited_by_count":20,"area":null},{"id":"2101.09868","title":"CPT: Efficient Deep Neural Network Training via Cyclic Precision","created":"2021-01-25","cited_by_count":19,"area":null},{"id":"2206.00843","title":"DepthShrinker: A New Compression Paradigm Towards Boosting Real-Hardware Efficiency of Compact Neural Networks","created":"2022-06-02","cited_by_count":19,"area":null},{"id":"2303.07129","title":"AdaptiveNet: Post-deployment Neural Architecture Adaptation for Diverse Edge Environments","created":"2023-03-13","cited_by_count":19,"area":null},{"id":"2311.15436","title":"Learning to Skip for Language Modeling","created":"2023-11-26","cited_by_count":19,"area":null},{"id":"2104.00749","title":"Anytime Dense Prediction with Confidence Adaptivity","created":"2021-04-01","cited_by_count":18,"area":null},{"id":"2106.03527","title":"Multi-Exit Semantic Segmentation Networks","created":"2021-06-07","cited_by_count":18,"area":null},{"id":"2205.06187","title":"Efficient Deep Visual and Inertial Odometry with Adaptive Visual Modality Selection","created":"2022-05-12","cited_by_count":18,"area":null},{"id":"2205.11269","title":"Dynamic Split Computing for Efficient Deep Edge Intelligence","created":"2022-05-23","cited_by_count":18,"area":null},{"id":"2105.04104","title":"AppealNet: An Efficient and Highly-Accurate Edge/Cloud Collaborative Architecture for DNN Inference","created":"2021-05-10","cited_by_count":17,"area":null},{"id":"2209.01425","title":"Dynamic Spatio-Temporal Specialization Learning for Fine-Grained Action Recognition","created":"2022-09-03","cited_by_count":17,"area":null},{"id":"2210.05370","title":"DeepPerform: An Efficient Approach for Performance Testing of Resource-Constrained Neural Networks","created":"2022-10-10","cited_by_count":17,"area":null},{"id":"2302.04453","title":"Data Quality-aware Mixed-precision Quantization via Hybrid Reinforcement Learning","created":"2023-02-09","cited_by_count":17,"area":null},{"id":"2210.03696","title":"LLMEffiChecker: Understanding and Testing Efficiency Degradation of Large Language Models","created":"2022-10-07","cited_by_count":16,"area":null},{"id":"2102.07887","title":"VA-RED$^2$: Video Adaptive Redundancy Reduction","created":"2021-02-15","cited_by_count":15,"area":null},{"id":"2304.05552","title":"DynamicDet: A Unified Dynamic Architecture for Object Detection","created":"2023-04-12","cited_by_count":15,"area":null},{"id":"2306.03445","title":"MetaGait: Learning to Learn an Omni Sample Adaptive Representation for Gait Recognition","created":"2023-06-06","cited_by_count":15,"area":null},{"id":"2306.04451","title":"Referring Expression Comprehension Using Language Adaptive Inference","created":"2023-06-06","cited_by_count":15,"area":null},{"id":"2306.16060","title":"Dynamic Path-Controllable Deep Unfolding Network for Compressive Sensing","created":"2023-06-28","cited_by_count":15,"area":null},{"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":"2206.07311","title":"Can pruning improve certified robustness of neural networks?","created":"2022-06-15","cited_by_count":14,"area":null},{"id":"2303.01980","title":"Towards energy-efficient Deep Learning: An overview of energy-efficient approaches along the Deep Learning Lifecycle","created":"2023-02-05","cited_by_count":14,"area":null},{"id":"2308.13998","title":"Computation-efficient Deep Learning for Computer Vision: A Survey","created":"2023-08-27","cited_by_count":14,"area":null},{"id":"2308.13494","title":"Eventful Transformers: Leveraging Temporal Redundancy in Vision Transformers","created":"2023-08-25","cited_by_count":13,"area":null},{"id":"2401.01543","title":"Retraining-free Model Quantization via One-Shot Weight-Coupling Learning","created":"2024-01-03","cited_by_count":13,"area":null},{"id":"2402.00033","title":"LF-ViT: Reducing Spatial Redundancy in Vision Transformer for Efficient Image Recognition","created":"2024-01-08","cited_by_count":13,"area":null},{"id":"2412.10739","title":"DSRC: Learning Density-insensitive and Semantic-aware Collaborative Representation against Corruptions","created":"2024-12-14","cited_by_count":13,"area":null},{"id":"2606.00198","title":"BAGEN: Are LLM Agents Budget-Aware?","created":"2026-05-29","cited_by_count":13,"area":"agents"},{"id":"2112.11542","title":"MIA-Former: Efficient and Robust Vision Transformers via Multi-grained Input-Adaptation","created":"2021-12-21","cited_by_count":12,"area":null},{"id":"2503.10905","title":"Learning to Inference Adaptively for Multimodal Large Language Models","created":"2025-03-13","cited_by_count":12,"area":"lr"},{"id":"2407.06677","title":"Mixture-of-Modules: Reinventing Transformers as Dynamic Assemblies of Modules","created":"2024-07-09","cited_by_count":11,"area":null},{"id":"2506.04179","title":"SkipGPT: Dynamic Layer Pruning Reinvented with Token Awareness and Module Decoupling","created":"2025-06-04","cited_by_count":11,"area":null},{"id":"2106.00261","title":"Exploring Dynamic Selection of Branch Expansion Orders for Code Generation","created":"2021-06-01","cited_by_count":10,"area":null},{"id":"2206.07839","title":"Linearity Grafting: Relaxed Neuron Pruning Helps Certifiable Robustness","created":"2022-06-15","cited_by_count":10,"area":null},{"id":"2302.04464","title":"Towards Fairer and More Efficient Federated Learning via Multidimensional Personalized Edge Models","created":"2023-02-09","cited_by_count":10,"area":null},{"id":"2303.07624","title":"I3D: Transformer architectures with input-dependent dynamic depth for speech recognition","created":"2023-03-14","cited_by_count":10,"area":null},{"id":"2303.07868","title":"DynaMask: Dynamic Mask Selection for Instance Segmentation","created":"2023-03-14","cited_by_count":10,"area":null}],"next_cursor":"100"}