{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/arxiv-2606-25285","title":"EPTS: Elastic Post-Training Sparsity for Efficient Large Language Model Compression","arxiv_id":"2606.25285","date":"2026-06-24","proceeding":null,"authors":["Ke Xu","Jiaqi Wan","Wenhao Hu","Han Pu","Xiaoyun Wang"],"abstract":"Post-Training Sparsity (PTS) has emerged as a crucial paradigm for compressing Large Language Models to facilitate efficient deployment on resource-constrained devices. However, existing PTS methodologies are typically confined to Single-Sparsity optimization, necessitating a separate, time-consuming optimization session for each specific sparsity level. This rigid paradigm significantly hinders flexible deployment across diverse hardware scenarios, as adapting to a new sparsity requirement mandates a complete re-optimization process. To address these limitations, we propose Elastic Post-Training Sparsity (EPTS), a unified Multi-Sparsity framework that produces a single elastic model capable of maintaining robust performance across diverse sparsity configurations through a one-shot optimization process. Specifically, we design a Multi-Sparsity Hierarchy LoRA (MS-HiLoRA) mechanism that facilitates knowledge inheritance from low- to high-sparsity groups, effectively mitigating the competition for parameter reconstruction. Furthermore, we introduce a Multi-Sparsity Feature Mixer (MSFM), which significantly enhances the model's adaptability to pruning perturbations by dynamically fusing feature representations of varying sparsity granularities. Extensive experiments on LLaMA and OPT families demonstrate that EPTS achieves competitive performance compared to state-of-the-art methods like SparseGPT and Wanda, while offering significant efficiency gains by enabling multi-scenario deployment from a single optimization. our source code is available at https://github.com/xuke225/EPTS.","url_abs":"https://arxiv.org/abs/2606.25285","url_pdf":"https://arxiv.org/pdf/2606.25285","source":{"archive":null,"snapshot":"2025-07-28","note":"not in the Papers with Code archive (frozen at the snapshot)","row_kind":"graph","title_abstract_authors_date":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)"},"code_links":[],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2606.25285","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2606.25285"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"mentioned_in_github":null,"is_official":null,"provenance":"deterministic:regex_extraction","mentioned_in_paper":null,"url":"https://github.com/xuke225/EPTS","reach":null}],"summary":{"ran_fixture":2,"ran":7},"by_repo_kind":{"found_in_text":{"samples":9,"ran":9,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":9,"samples":[{"code_sha256_prefix":"b0f936d4d6ae3b8c","entry":"accuracy","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/imagenet_utils.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/imagenet_utils.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b0f936d4d6ae3b8c"}},{"code_sha256_prefix":"fbab3eee6f932445","entry":"build_imagenet_data","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/imagenet.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/imagenet.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"fbab3eee6f932445"}},{"code_sha256_prefix":"2db18966f7256dba","entry":"count_nonzero_parameters","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"inference.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/inference.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2db18966f7256dba"}},{"code_sha256_prefix":"8b8839a8e847a6c5","entry":"generate_llama_uniform_rates","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/reconstruction.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/reconstruction.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8b8839a8e847a6c5"}},{"code_sha256_prefix":"69638fd064d7f1e0","entry":"get_batch_outputs","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/reconstruction.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/reconstruction.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"69638fd064d7f1e0"}},{"code_sha256_prefix":"0809140f7f488d09","entry":"get_llm","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"inference.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/inference.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"0809140f7f488d09"}},{"code_sha256_prefix":"b52566618ab18da7","entry":"get_llm","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"precomputer_wanda_importance.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/precomputer_wanda_importance.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b52566618ab18da7"}},{"code_sha256_prefix":"b6712d04d930c9bb","entry":"load_data","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/imagenet_utils.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/imagenet_utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b6712d04d930c9bb"}},{"code_sha256_prefix":"8fcf6c2e205ea12b","entry":"lp_loss","repo":"xuke225/EPTS","repo_kind":"found_in_text","path":"utils/reconstruction.py","file_url":"https://github.com/xuke225/EPTS/blob/HEAD/utils/reconstruction.py","link_basis":"plan_row","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8fcf6c2e205ea12b"}}]},"arxiv_metadata":{"licence":"arXiv metadata, CC0 1.0 (https://info.arxiv.org/help/license)","fields":["title","abstract","authors","date"],"primary_category":"cs.LG","source":"arxiv_2026.jsonl"},"syntology_extracted_results":null}