{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2026,7,14]],"date-time":"2026-07-14T15:06:41Z","timestamp":1784041601759,"version":"3.55.0"},"reference-count":237,"publisher":"Institute of Electrical and Electronics Engineers (IEEE)","issue":"12","license":[{"start":{"date-parts":[[2024,12,1]],"date-time":"2024-12-01T00:00:00Z","timestamp":1733011200000},"content-version":"vor","delay-in-days":0,"URL":"https:\/\/ieeexplore.ieee.org\/Xplorehelp\/downloads\/license-information\/IEEE.html"},{"start":{"date-parts":[[2024,12,1]],"date-time":"2024-12-01T00:00:00Z","timestamp":1733011200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-029"},{"start":{"date-parts":[[2024,12,1]],"date-time":"2024-12-01T00:00:00Z","timestamp":1733011200000},"content-version":"stm-asf","delay-in-days":0,"URL":"https:\/\/doi.org\/10.15223\/policy-037"}],"funder":[{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["62306084"],"award-info":[{"award-number":["62306084"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"DOI":"10.13039\/501100001809","name":"National Natural Science Foundation of China","doi-asserted-by":"publisher","award":["U23B2051"],"award-info":[{"award-number":["U23B2051"]}],"id":[{"id":"10.13039\/501100001809","id-type":"DOI","asserted-by":"publisher"}]},{"name":"Shenzhen College Stability Support Plan","award":["GXWD20231128102243003"],"award-info":[{"award-number":["GXWD20231128102243003"]}]}],"content-domain":{"domain":[],"crossmark-restriction":false},"short-container-title":["IEEE Trans. Pattern Anal. Mach. Intell."],"published-print":{"date-parts":[[2024,12]]},"DOI":"10.1109\/tpami.2024.3447085","type":"journal-article","created":{"date-parts":[[2024,8,21]],"date-time":"2024-08-21T23:25:34Z","timestamp":1724282734000},"page":"10558-10578","source":"Crossref","is-referenced-by-count":417,"title":["A Survey on Deep Neural Network Pruning: Taxonomy, Comparison, Analysis, and Recommendations"],"prefix":"10.1109","volume":"46","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-4766-8935","authenticated-orcid":false,"given":"Hongrong","family":"Cheng","sequence":"first","affiliation":[{"name":"University of Adelaide, Adelaide, SA, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-1262-4174","authenticated-orcid":false,"given":"Miao","family":"Zhang","sequence":"additional","affiliation":[{"name":"Harbin Institute of Technology, Shenzhen, China"}],"role":[{"vocabulary":"crossref","role":"author"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-9126-2107","authenticated-orcid":false,"given":"Javen Qinfeng","family":"Shi","sequence":"additional","affiliation":[{"name":"University of Adelaide, Adelaide, SA, Australia"}],"role":[{"vocabulary":"crossref","role":"author"}]}],"member":"263","reference":[{"key":"ref1","first-page":"1","article-title":"Very deep convolutional networks for large-scale image recognition","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Simonyan"},{"key":"ref2","article-title":"An image is worth 16x16 words: Transformers for image recognition at scale","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Dosovitskiy"},{"key":"ref3","doi-asserted-by":"publisher","DOI":"10.48550\/arXiv.1810.04805"},{"key":"ref4","article-title":"PaLM: Scaling language modeling with pathways","author":"Chowdhery","year":"2022"},{"key":"ref5","article-title":"Towards efficient end-to-end speech recognition with biologically-inspired neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Bohnstingl"},{"key":"ref6","article-title":"Sparks of large audio models: A survey and outlook","author":"Latif","year":"2023"},{"key":"ref7","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52733.2024.02520"},{"key":"ref8","article-title":"Sora: A review on background, technology, limitations, and opportunities of large vision models","author":"Liu","year":"2024"},{"key":"ref9","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2016.90"},{"key":"ref10","article-title":"Drawing early-bird tickets: Towards more efficient training of deep networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"You"},{"key":"ref11","article-title":"AI-generated content AIGC: A survey","author":"Wu","year":"2023"},{"key":"ref12","first-page":"1135","article-title":"Learning both weights and connections for efficient neural network","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Han"},{"key":"ref13","article-title":"Deep compression: Compressing deep neural networks with pruning, trained quantization and Huffman coding","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Han"},{"key":"ref14","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2017.205"},{"key":"ref15","first-page":"2130","article-title":"Gate Decorator: Global filter pruning method for accelerating deep convolutional neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"You"},{"key":"ref16","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.541"},{"key":"ref17","article-title":"SparseGPT: Massive language models can be accurately pruned in one-shot","author":"Frantar","year":"2023"},{"key":"ref18","article-title":"HYDRA: Pruning adversarially robust neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Sehwag"},{"key":"ref19","article-title":"SliceGPT: Compress large language models by deleting rows and columns","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Ashkboos"},{"key":"ref20","first-page":"21702","article-title":"LLM-Pruner: On the structural pruning of large language models","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Ma"},{"key":"ref21","first-page":"1269","article-title":"Exploiting linear structure within convolutional networks for efficient evaluation","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Denton"},{"key":"ref22","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2018.2873305"},{"key":"ref23","article-title":"QLoRA: Efficient finetuning of quantized LLMs","author":"Dettmers","year":"2023"},{"key":"ref24","article-title":"OmniQuant: Omnidirectionally calibrated quantization for large language models","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Shao"},{"key":"ref25","article-title":"MiniLLM: Knowledge distillation of large language models","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Gu"},{"key":"ref26","article-title":"A survey on knowledge distillation of large language models","author":"Xu","year":"2024"},{"key":"ref27","article-title":"DARTS: Differentiable architecture search","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu"},{"key":"ref28","first-page":"12557","article-title":"iDARTS: Differentiable architecture search with stochastic implicit gradients","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Zhang"},{"key":"ref29","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01352"},{"key":"ref30","article-title":"MobileVLM V2: Faster and stronger baseline for vision language model","author":"Chu","year":"2024"},{"key":"ref31","first-page":"177","article-title":"Comparing biases for minimal network construction with back-propagation","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Hanson"},{"key":"ref32","article-title":"A survey on deep neural network compression: Challenges, overview, and solutions","author":"Mishra","year":"2020"},{"key":"ref33","doi-asserted-by":"publisher","DOI":"10.3390\/electronics11060945"},{"key":"ref34","article-title":"A comprehensive survey of compression algorithms for language models","author":"Park","year":"2024"},{"key":"ref35","article-title":"A survey of resource-efficient LLM and multimodal foundation models","author":"Xu","year":"2024"},{"key":"ref36","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2022\/786"},{"key":"ref37","article-title":"Structured pruning for deep convolutional neural networks: A survey","author":"He","year":"2023"},{"key":"ref38","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01544"},{"key":"ref39","article-title":"Structural pruning for diffusion models","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Fang"},{"key":"ref40","article-title":"A survey on model compression for large language models","author":"Zhu","year":"2023"},{"key":"ref41","doi-asserted-by":"publisher","DOI":"10.1109\/72.248452"},{"key":"ref42","first-page":"129","article-title":"What is the state of neural network pruning?","volume-title":"Proc. Mach. Learn. Syst.","author":"Blalock"},{"key":"ref43","article-title":"An survey of neural network compression","author":"O\u2019Neill","year":"2020"},{"key":"ref44","article-title":"Pruning algorithms to accelerate convolutional neural networks for edge applications: A survey","author":"Liu","year":"2020"},{"key":"ref45","doi-asserted-by":"publisher","DOI":"10.1007\/s10462-020-09816-7"},{"key":"ref46","article-title":"Sparsity in deep learning: Pruning and growth for efficient inference and training in neural networks","volume":"22","author":"Hoefler","year":"2021","journal-title":"J. Mach. Learn. Res."},{"key":"ref47","article-title":"Model compression and efficient inference for large language models: A survey","author":"Wang","year":"2024"},{"key":"ref48","article-title":"The lottery ticket hypothesis: Finding sparse, trainable neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Frankle"},{"key":"ref49","article-title":"A simple and effective pruning approach for large language models","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Sun"},{"key":"ref50","first-page":"6377","article-title":"Pruning neural networks without any data by iteratively conserving synaptic flow","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Tanaka"},{"key":"ref51","article-title":"Pruning filter in filter","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Meng"},{"key":"ref52","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58601-0_37"},{"key":"ref53","article-title":"Trainability preserving neural structured pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang"},{"key":"ref54","article-title":"SNIP: Single-shot network pruning based on connection sensitivity","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lee"},{"key":"ref55","first-page":"20390","article-title":"Sanity-checking pruning methods: Random tickets can win the jackpot","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Su"},{"key":"ref56","article-title":"Picking winning tickets before training by preserving gradient flow","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang"},{"key":"ref57","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.298"},{"key":"ref58","doi-asserted-by":"publisher","DOI":"10.5555\/3157096.3157329"},{"key":"ref59","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01270-0_19"},{"key":"ref60","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.01779"},{"key":"ref61","article-title":"Dynamic sparse training: Find efficient sparse network from scratch with trainable masked layers","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu"},{"key":"ref62","first-page":"7021","article-title":"Group fisher pruning for practical network compression","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Liu"},{"key":"ref63","article-title":"Pruning filters for efficient ConvNets","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Li"},{"key":"ref64","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00498"},{"key":"ref65","article-title":"ShortGPT: Layers in large language models are more redundant than you expect","author":"Men","year":"2024"},{"key":"ref66","doi-asserted-by":"publisher","DOI":"10.48550\/ARXIV.1706.03762"},{"key":"ref67","article-title":"SOSP: Efficiently capturing global correlations by second-order structured pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Nonnenmacher"},{"key":"ref68","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01403"},{"key":"ref69","article-title":"The unreasonable effectiveness of random pruning: Return of the most naive baseline for sparse training","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu"},{"key":"ref70","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v34i07.6910"},{"key":"ref71","article-title":"Dual lottery ticket hypothesis","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Bai"},{"key":"ref72","article-title":"A signal propagation perspective for pruning neural networks at initialization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lee"},{"key":"ref73","article-title":"Exact solutions to the nonlinear dynamics of learning in deep linear neural networks","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Saxe"},{"key":"ref74","article-title":"A unified paths perspective for pruning at initialization","author":"Gebhart","year":"2021"},{"key":"ref75","first-page":"8580","article-title":"Neural tangent kernel: Convergence and generalization in neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Jacot"},{"key":"ref76","article-title":"Pruning neural networks at initialization: Why are we missing the mark?","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Frankle"},{"key":"ref77","first-page":"5365","article-title":"Weight agnostic neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Gaier"},{"key":"ref78","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01191"},{"key":"ref79","article-title":"Revisiting pruning at initialization through the lens of Ramanujan graph","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hoang"},{"key":"ref80","article-title":"No free prune: Information-theoretic barriers to pruning at initialization","author":"Kumar","year":"2024"},{"key":"ref81","first-page":"2943","article-title":"Rigging the lottery: Making all tickets winners","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Evci"},{"key":"ref82","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00289"},{"key":"ref83","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2018\/309"},{"key":"ref84","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2019.00339"},{"key":"ref85","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58580-8_35"},{"key":"ref86","article-title":"PDP: Parameter-free differentiable pruning is all you need","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Cho"},{"key":"ref87","doi-asserted-by":"publisher","DOI":"10.1111\/j.1467-9868.2005.00532.x"},{"key":"ref88","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00171"},{"key":"ref89","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00410"},{"key":"ref90","doi-asserted-by":"publisher","DOI":"10.1561\/9781601984616"},{"key":"ref91","doi-asserted-by":"publisher","DOI":"10.1038\/s41467-018-04316-3"},{"key":"ref92","first-page":"9908","article-title":"Sparse training via boosting pruning plasticity with neuroregeneration","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Liu"},{"key":"ref93","article-title":"Deep ensembling with no overhead for either training or testing: The all-round blessings of dynamic sparsity","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu"},{"key":"ref94","doi-asserted-by":"publisher","DOI":"10.1109\/TC.2019.2914438"},{"key":"ref95","first-page":"4646","article-title":"Parameter efficient training of deep convolutional neural networks by dynamic sparse reparameterization","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Mostafa"},{"key":"ref96","article-title":"Dynamic model pruning with feedback","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lin"},{"key":"ref97","article-title":"Sparse networks from scratch: Faster training without losing performance","author":"Dettmers","year":"2019"},{"key":"ref98","doi-asserted-by":"publisher","DOI":"10.24963\/ijcai.2022\/477"},{"key":"ref99","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v36i6.20611"},{"key":"ref100","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00447"},{"key":"ref101","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2008.4587747"},{"key":"ref102","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00161"},{"key":"ref103","doi-asserted-by":"publisher","DOI":"10.1155\/2022\/7775419"},{"key":"ref104","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52729.2023.02333"},{"key":"ref105","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v32i1.12262"},{"key":"ref106","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01152"},{"key":"ref107","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2019.2906563"},{"key":"ref108","article-title":"Neural pruning via growing regularization","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Wang"},{"key":"ref109","article-title":"Compressing large language models by streamlining the unimportant layer","author":"Chen","year":"2024"},{"key":"ref110","article-title":"Rethinking the value of network pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Liu"},{"key":"ref111","article-title":"Comparing rewinding and fine-tuning in neural network pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Renda"},{"key":"ref112","article-title":"A survey of lottery ticket hypothesis","author":"Liu","year":"2024"},{"key":"ref113","article-title":"Multi-prize lottery ticket hypothesis: Finding a accurate binary neural networks by pruning a randomly weighted network","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Diffenderfer"},{"key":"ref114","article-title":"GANs can play lottery tickets too","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chen"},{"key":"ref115","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v35i9.16980"},{"key":"ref116","article-title":"Rethinking graph lottery tickets: Graph sparsity matters","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hui"},{"key":"ref117","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.emnlp-main.259"},{"key":"ref118","first-page":"6682","article-title":"Proving the lottery ticket hypothesis: Pruning is all you need","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Malach"},{"key":"ref119","article-title":"Logarithmic pruning is all you need","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Orseau"},{"key":"ref120","first-page":"4933","article-title":"One ticket to win them all: Generalizing lottery ticket initializations across datasets and optimizers","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Morcos"},{"key":"ref121","article-title":"Sparse transfer learning via winning lottery tickets","author":"Mehta","year":"2019"},{"key":"ref122","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v36i1.19945"},{"key":"ref123","first-page":"1695","article-title":"A unified lottery ticket hypothesis for graph neural networks","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Chen"},{"key":"ref124","first-page":"30196","article-title":"Validating the lottery ticket hypothesis with inertial manifold theory","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Zhang"},{"key":"ref125","article-title":"Why lottery ticket wins? a theoretical perspective of sample complexity on pruned neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Zhang"},{"key":"ref126","article-title":"Unmasking the lottery ticket hypothesis-what\u2019s encoded in a winning ticket\u2019s mask","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Paul"},{"key":"ref127","first-page":"12749","article-title":"Sanity checks for lottery tickets: Does your winning ticket really win the jackpot?","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Ma"},{"key":"ref128","first-page":"7011","article-title":"Lottery ticket preserves weight correlation: Is it desirable or not?","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Liu"},{"key":"ref129","article-title":"LoRA: Low-rank adaptation of large language models","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Hu"},{"key":"ref130","first-page":"31292","article-title":"UPop: Unified and progressive pruning for compressing vision-language transformers","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Shi"},{"key":"ref131","first-page":"26326","article-title":"Gradient-free structured pruning with unlabeled data","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Nova"},{"key":"ref132","article-title":"Everybody prune now: Structured pruning of LLMs with only forward passes","author":"Dery","year":"2024"},{"key":"ref133","article-title":"Shortened LLaMA: A simple depth pruning for large language models","author":"Kim","year":"2024"},{"key":"ref134","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV.2017.155"},{"key":"ref135","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00290"},{"key":"ref136","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.acl-long.107"},{"key":"ref137","article-title":"Sheared LLaMA: Accelerating language model pre-training via structured pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Xia"},{"key":"ref138","doi-asserted-by":"publisher","DOI":"10.1111\/j.2517-6161.1996.tb02080.x"},{"key":"ref139","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.01146"},{"key":"ref140","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00915"},{"key":"ref141","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.acl-long.171"},{"key":"ref142","first-page":"14014","article-title":"Are sixteen heads really better than one?","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Michel"},{"key":"ref143","first-page":"3259","article-title":"Linear mode connectivity and the lottery ticket hypothesis","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Frankle"},{"key":"ref144","article-title":"A fast post-training pruning framework for transformers","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Kwon"},{"key":"ref145","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i10.28960"},{"key":"ref146","doi-asserted-by":"publisher","DOI":"10.1109\/TPAMI.2018.2878258"},{"key":"ref147","first-page":"1886","article-title":"Channel gating neural networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Hua"},{"key":"ref148","article-title":"Dynamic channel pruning: Feature boosting and suppression","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Gao"},{"key":"ref149","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01213"},{"key":"ref150","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01194"},{"key":"ref151","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2006.100"},{"key":"ref152","doi-asserted-by":"publisher","DOI":"10.1109\/TCAD.2023.3273992"},{"key":"ref153","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/K16-1029"},{"key":"ref154","article-title":"A gradient flow framework for analyzing network pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lubana"},{"key":"ref155","article-title":"Towards compact and robust deep neural networks","author":"Sehwag","year":"2019"},{"key":"ref156","article-title":"Pruning convolutional neural networks for resource efficient inference","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Molchanov"},{"key":"ref157","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.00637"},{"key":"ref158","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v36i3.20222"},{"key":"ref159","first-page":"9075","article-title":"On the predictability of pruning across scales","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Rosenfeld"},{"key":"ref160","article-title":"Layer-adaptive sparsity for the magnitude-based pruning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Lee"},{"key":"ref161","first-page":"598","article-title":"Optimal brain damage","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"LeCun"},{"key":"ref162","article-title":"What matters in the structured pruning of generative language models?","author":"Santacroce","year":"2023"},{"key":"ref163","article-title":"The LLM surgeon","author":"van der Ouderaa","year":"2024"},{"key":"ref164","first-page":"164","article-title":"Second order derivatives for network pruning: Optimal brain surgeon","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Hassibi"},{"key":"ref165","first-page":"6566","article-title":"EigenDamage: Structured pruning in the Kronecker-Factored eigenbasis","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Wang"},{"key":"ref166","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2022.emnlp-main.279"},{"key":"ref167","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58598-3_36"},{"key":"ref168","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/P19-1580"},{"key":"ref169","doi-asserted-by":"publisher","DOI":"10.1007\/s10489-021-02802-8"},{"key":"ref170","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01234-2_48"},{"key":"ref171","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00257"},{"key":"ref172","article-title":"Single shot structured pruning before training","author":"Amersfoort","year":"2020"},{"key":"ref173","article-title":"Pointer sentinel mixture models","author":"Merity","year":"2016"},{"key":"ref174","article-title":"OPT: Open pre-trained transformer language models","author":"Zhang","year":"2022"},{"key":"ref175","first-page":"97","article-title":"Learning methods for generic object recognition with invariance to pose and lighting","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit.","author":"Huang"},{"key":"ref176","article-title":"Learning multiple layers of features from tiny images","author":"Krizhevsky","year":"2009"},{"key":"ref177","article-title":"LLaMA: Open and efficient foundation language models","author":"Touvron","year":"2023"},{"key":"ref178","doi-asserted-by":"publisher","DOI":"10.21236\/ADA273556"},{"key":"ref179","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2019.00291"},{"key":"ref180","doi-asserted-by":"publisher","DOI":"10.1007\/s11263-015-0816-y"},{"key":"ref181","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2024.findings-acl.178"},{"key":"ref182","article-title":"Why is pruning at initialization immune to reinitializing and shuffling?","author":"Singh","year":"2021"},{"key":"ref183","article-title":"Train-by-Reconnect: Decoupling locations of weights from their values","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Qiu"},{"key":"ref184","first-page":"1","article-title":"How much pre-training is enough to discover a good subnetwork?","volume":"2024","author":"Wolfe","year":"2024","journal-title":"Trans. Mach. Learn. Res."},{"key":"ref185","article-title":"Never train from scratch: Fair comparison of long-sequence models requires data-driven priors","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Amos"},{"key":"ref186","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00159"},{"key":"ref187","article-title":"ChipNet: Budget-aware pruning with heaviside continuous approximations","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Tiwari"},{"key":"ref188","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2021.findings-acl.300"},{"key":"ref189","article-title":"SparseLLM: Towards global pruning for pre-trained language models","author":"Bai","year":"2024"},{"key":"ref190","article-title":"Network pruning that matters: A case study on pretraing variants","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Le"},{"key":"ref191","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-01237-3_12"},{"key":"ref192","first-page":"10820","article-title":"Good subnetworks provably exist: Pruning via greedy forward selection","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Ye"},{"key":"ref193","first-page":"10347","article-title":"Training data-efficient image transformers & distillation through attention","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Touvron"},{"key":"ref194","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2009.191"},{"key":"ref195","article-title":"The lottery ticket hypothesis for pre-trained BERT networks","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Chen"},{"key":"ref196","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01604"},{"key":"ref197","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.01195"},{"key":"ref198","first-page":"26609","article-title":"The elastic lottery ticket hypothesis","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Chen"},{"key":"ref199","doi-asserted-by":"publisher","DOI":"10.1109\/ICCV48922.2021.00528"},{"key":"ref200","article-title":"Movement pruning: Adaptive sparsity by fine-tuning","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Sanh"},{"key":"ref201","article-title":"Are all layers created equal?","volume":"23","author":"Zhang","year":"2022","journal-title":"J. Mach. Learn. Res."},{"key":"ref202","doi-asserted-by":"publisher","DOI":"10.1609\/aaai.v38i11.29108"},{"key":"ref203","article-title":"Outlier weighed layerwise sparsity (OWL): A missing secret sauce for pruning LLMs to high sparsity","author":"Yin","year":"2024"},{"key":"ref204","article-title":"Why lift so heavy? Slimming large language models by cutting off the layers","author":"Yuan","year":"2024"},{"key":"ref205","article-title":"LaCo: Large language model pruning via layer collapse","author":"Yang","year":"2024"},{"key":"ref206","doi-asserted-by":"publisher","DOI":"10.1109\/BIOCAS.2018.8584839"},{"key":"ref207","doi-asserted-by":"publisher","DOI":"10.1109\/TKDE.2021.3090866"},{"key":"ref208","article-title":"Pruning convolutional neural networks with self-supervision","author":"Caron","year":"2020"},{"key":"ref209","doi-asserted-by":"publisher","DOI":"10.1109\/CVPRW56347.2022.00298"},{"key":"ref210","doi-asserted-by":"publisher","DOI":"10.5555\/3524938.3525087"},{"key":"ref211","article-title":"Improved baselines with momentum contrastive learning","author":"Chen","year":"2020"},{"key":"ref212","first-page":"21256","article-title":"PARP: Prune, adjust and re-prune for self-supervised speech recognition","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Jeff Lai"},{"key":"ref213","article-title":"Distilling the knowledge in a neural network","author":"Hinton","year":"2015"},{"key":"ref214","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR.2018.00821"},{"key":"ref215","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.00804"},{"key":"ref216","first-page":"20336","article-title":"LoSparse: Structured compression of large language models based on low-rank and sparse approximation","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Li"},{"key":"ref217","first-page":"759","article-title":"Network pruning via transformable architecture search","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Dong"},{"key":"ref218","article-title":"Structural pruning of large language models via neural architecture search","volume-title":"Proc. Int. Conf. Autom. Mach. Learn.","author":"Klein"},{"key":"ref219","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01198"},{"key":"ref220","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-031-20083-0_8"},{"key":"ref221","doi-asserted-by":"publisher","DOI":"10.1007\/s40747-020-00248-y"},{"key":"ref222","article-title":"Long live the lottery: The existence of winning tickets in lifelong learning","volume-title":"Proc. Int. Conf. Learn. Representations","author":"Chen"},{"key":"ref223","first-page":"18309","article-title":"Advancing model pruning via bi-level optimization","volume-title":"Proc. Int. Conf. Neural Inf. Process. Syst.","author":"Zhang"},{"key":"ref224","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00593"},{"key":"ref225","article-title":"Large multimodal model compression via efficient pruning and distillation at AntGroup","author":"Wang","year":"2023"},{"key":"ref226","doi-asserted-by":"publisher","DOI":"10.18653\/v1\/2020.coling-main.287"},{"key":"ref227","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR46437.2021.01004"},{"key":"ref228","doi-asserted-by":"publisher","DOI":"10.1007\/978-3-030-58548-8_4"},{"key":"ref229","article-title":"Train large, then compress: Rethinking model size for efficient training and inference of transformers","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Li"},{"key":"ref230","first-page":"15751","article-title":"ParameterNet: Parameters are all you need","volume-title":"Proc. IEEE Conf. Comput. Vis. Pattern Recognit.","author":"Han"},{"key":"ref231","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR42600.2020.01104"},{"key":"ref232","doi-asserted-by":"publisher","DOI":"10.1109\/CVPR52688.2022.00021"},{"key":"ref233","article-title":"Studying the impact of magnitude pruning on contrastive learning methods","volume-title":"Proc. Int. Conf. Mach. Learn.","author":"Corti"},{"key":"ref234","doi-asserted-by":"publisher","DOI":"10.1109\/TNNLS.2022.3166101"},{"key":"ref235","doi-asserted-by":"publisher","DOI":"10.3390\/s23020824"},{"key":"ref236","article-title":"GPT-4 technical report","author":"Achiam","year":"2023"},{"key":"ref237","article-title":"Why is the state of neural network pruning so confusing? On the fairness, comparison setup, and trainability in network pruning","author":"Wang","year":"2023"}],"container-title":["IEEE Transactions on Pattern Analysis and Machine Intelligence"],"original-title":[],"link":[{"URL":"http:\/\/xplorestaging.ieee.org\/ielx8\/34\/10746266\/10643325.pdf?arnumber=10643325","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,11,27]],"date-time":"2024-11-27T00:31:43Z","timestamp":1732667503000},"score":1,"resource":{"primary":{"URL":"https:\/\/ieeexplore.ieee.org\/document\/10643325\/"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024,12]]},"references-count":237,"journal-issue":{"issue":"12"},"URL":"https:\/\/doi.org\/10.1109\/tpami.2024.3447085","relation":{},"ISSN":["0162-8828","2160-9292","1939-3539"],"issn-type":[{"value":"0162-8828","type":"print"},{"value":"2160-9292","type":"electronic"},{"value":"1939-3539","type":"electronic"}],"subject":[],"published":{"date-parts":[[2024,12]]}}}