{"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-2605-28819","title":"PEFT-Arena: Understanding Parameter-Efficient Finetuning from a Stability-Plasticity Perspective","arxiv_id":"2605.28819","date":"2026-05-27","proceeding":null,"authors":["Yangyi Huang","Ruotian Peng","Zeju Qiu","Jiale Kang","Yandong Wen","Bernhard Schölkopf","Weiyang Liu"],"abstract":"Parameter-efficient finetuning (PEFT) has become the standard approach for adapting large language models, yet evaluations largely emphasize downstream accuracy while overlooking the retention of pretrained capabilities. We argue that PEFT should be assessed through the stability-plasticity dilemma: the trade-off between target-task adaptation and resistance to forgetting. We introduce PEFT-Arena, a benchmark that jointly measures downstream performance and general capability retention. Across methods, we find distinct stability-plasticity profiles; under comparable parameter budgets, orthogonal finetuning achieves the most favorable Pareto frontier. To explain these differences, we analyze PEFT updates from two geometric perspectives. In weight space, spectral analysis reveals how parameterizations interact with the pretrained singular-value structure. In activation space, retention metrics show whether finetuning preserves or distorts general-capability representations, with forgetting linked to non-isometric representation distortion. Finally, an analysis shows that final SFT checkpoints often overshoot a better target-retention operating point. Inspired by this, we present case studies of a post-hoc improvement with path-wise rewinding.","url_abs":"https://arxiv.org/abs/2605.28819","url_pdf":"https://arxiv.org/pdf/2605.28819","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=2605.28819","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.28819"}},"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/Sphere-AI-Lab/PEFT-Arena","reach":null}],"summary":{"ran_fixture":4,"ran_honours":1},"by_repo_kind":{"found_in_text":{"samples":5,"ran":5,"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":0,"samples":[{"code_sha256_prefix":"578425d5cc41a94e","entry":"_compute_principal_basis","repo":"Sphere-AI-Lab/PEFT-Arena","repo_kind":"found_in_text","path":"train/peft_arena_verl/utils/keeplora.py","file_url":"https://github.com/Sphere-AI-Lab/PEFT-Arena/blob/HEAD/train/peft_arena_verl/utils/keeplora.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"578425d5cc41a94e"}},{"code_sha256_prefix":"88231bac2b16a62f","entry":"_safe_lowrank_q","repo":"Sphere-AI-Lab/PEFT-Arena","repo_kind":"found_in_text","path":"train/peft_arena_verl/utils/keeplora.py","file_url":"https://github.com/Sphere-AI-Lab/PEFT-Arena/blob/HEAD/train/peft_arena_verl/utils/keeplora.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"88231bac2b16a62f"}},{"code_sha256_prefix":"4f1583ebfcc92192","entry":"build_keeplora_factors","repo":"Sphere-AI-Lab/PEFT-Arena","repo_kind":"found_in_text","path":"train/peft_arena_verl/utils/keeplora.py","file_url":"https://github.com/Sphere-AI-Lab/PEFT-Arena/blob/HEAD/train/peft_arena_verl/utils/keeplora.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"4f1583ebfcc92192"}},{"code_sha256_prefix":"0ff3809222023c7e","entry":"initialize_keeplora_adapters","repo":"Sphere-AI-Lab/PEFT-Arena","repo_kind":"found_in_text","path":"train/peft_arena_verl/utils/keeplora.py","file_url":"https://github.com/Sphere-AI-Lab/PEFT-Arena/blob/HEAD/train/peft_arena_verl/utils/keeplora.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"0ff3809222023c7e"}},{"code_sha256_prefix":"a8ff4f381a2015f9","entry":"project_gradient_to_residual_subspace","repo":"Sphere-AI-Lab/PEFT-Arena","repo_kind":"found_in_text","path":"train/peft_arena_verl/utils/keeplora.py","file_url":"https://github.com/Sphere-AI-Lab/PEFT-Arena/blob/HEAD/train/peft_arena_verl/utils/keeplora.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"a8ff4f381a2015f9"}}]},"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}