{"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-2605-30992","title":"Eigenvectors of Experts are Training-free Non-collapsing Routers","arxiv_id":"2605.30992","date":"2026-05-29","proceeding":"ICML","authors":["Giang Do","Hung Le","Truyen Tran"],"abstract":"Sparse Mixture of Experts (SMoE) architectures improve the training efficiency of Large Language Models (LLMs) by routing input tokens to a selected subset of specialized experts. Despite their remarkable success, both training and inference in SMoE models suffer from the expert collapse issue (Chi et al., 2022), which degrades model performance. Prior studies primarily focus on improving the router; however, such methods rely on training from scratch or fine-tuning, which requires high computational and data-processing costs. Furthermore, we demonstrate that, despite these efforts, the issue persists when advancing well-pretrained SMoE models, as evidenced by both theoretical and empirical results. To fill that gap, we analyze the advanced SMoE models and observe that the eigenvectors of expert weight matrices encode rich semantic information, pointing to an effective alternative to conventional routing strategies. Building on this insight, we propose Singular Value Decomposition SMoE (SSMoE), a novel and training-free framework that leverages spectral properties of the expert weights to address the collapse issue and enhance model performance. Extensive experiments across diverse language and vision tasks, under both clean and corrupt data settings, demonstrate the strong generalization and robustness of SSMoE. Our findings highlight how a deeper understanding of model internals can guide the development of more effective SMoE architectures. Our implementation is publicly available at https://github.com/giangdip2410/SSMoE.","url_abs":"https://arxiv.org/abs/2605.30992","url_pdf":"https://arxiv.org/pdf/2605.30992","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=2605.30992","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2605.30992"}},"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/giangdip2410/SSMoE","reach":null}],"summary":{"ran_fixture":2,"ran_draft_wrong":5,"ran_violates":1,"ran":2,"unverified":1},"by_repo_kind":{"found_in_text":{"samples":11,"ran":10,"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":11,"samples":[{"code_sha256_prefix":"30d7eec482ebf6b1","entry":"repeat_kv","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":2,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"30d7eec482ebf6b1"}},{"code_sha256_prefix":"d61c483a3c2b3156","entry":"apply_rotary_pos_emb","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","link_basis":"plan_row","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":"d61c483a3c2b3156"}},{"code_sha256_prefix":"bac65c3dafaec040","entry":"apply_rotary_pos_emb","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_olmoe.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_olmoe.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"bac65c3dafaec040"}},{"code_sha256_prefix":"aecadcaa2ac48b1b","entry":"get_gpus_max_memory","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/eval_mteb.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/eval_mteb.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"aecadcaa2ac48b1b"}},{"code_sha256_prefix":"62866964865efbd3","entry":"gritlm_instruction_format","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/eval_mteb.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/eval_mteb.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"62866964865efbd3"}},{"code_sha256_prefix":"d0725ff0666ee959","entry":"load_balancing_loss_func","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_olmoe.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_olmoe.py","link_basis":"plan_row","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"d0725ff0666ee959"}},{"code_sha256_prefix":"e566aca3f6f9392e","entry":"load_balancing_loss_func","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_gritlm8x7b.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_gritlm8x7b.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":"e566aca3f6f9392e"}},{"code_sha256_prefix":"25b8549e7bd56c7d","entry":"load_balancing_loss_func","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Vision/CLIP-SSMoE/clipmoe/model_clipmoe.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Vision/CLIP-SSMoE/clipmoe/model_clipmoe.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":"25b8549e7bd56c7d"}},{"code_sha256_prefix":"b99eea6376d1e212","entry":"rotate_half","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/models/modeling_deepseek.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"b99eea6376d1e212"}},{"code_sha256_prefix":"a591f38d473d9770","entry":"zephyr_instruction_format","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Language/SSMoE-Embedding/eval_mteb.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Language/SSMoE-Embedding/eval_mteb.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a591f38d473d9770"}},{"code_sha256_prefix":"640bc7cbe2f90309","entry":"build_model","repo":"giangdip2410/SSMoE","repo_kind":"found_in_text","path":"SSMoE/Vision/CLIP-SSMoE/clipmoe/model_clipmoe.py","file_url":"https://github.com/giangdip2410/SSMoE/blob/HEAD/SSMoE/Vision/CLIP-SSMoE/clipmoe/model_clipmoe.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"640bc7cbe2f90309"}}]},"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}