{"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/dyad-a-descriptive-yet-abjuring-density","title":"DYAD: A Descriptive Yet Abjuring Density efficient approximation to linear neural network layers","arxiv_id":"2312.06881","date":"2023-12-11","proceeding":null,"authors":["Sarin Chandy","Varun Gangal","Yi Yang","Gabriel Maggiotti"],"abstract":"We devise, implement and performance-asses DYAD, a layer which can serve as a faster and more memory-efficient approximate replacement for linear layers, (nn.Linear() in Pytorch). These layers appear in common subcomponents, such as in the ff module of Transformers. DYAD is based on a bespoke near-sparse matrix structure which approximates the dense \"weight\" matrix W that matrix-multiplies the input in the typical realization of such a layer, a.k.a DENSE. Our alternative near-sparse matrix structure is decomposable to a sum of 2 matrices permutable to a block-sparse counterpart. These can be represented as 3D tensors, which in unison allow a faster execution of matrix multiplication with the mini-batched input matrix X compared to DENSE (O(rows(W ) x cols(W )) --> O( rows(W ) x cols(W ) # of blocks )). As the crux of our experiments, we pretrain both DYAD and DENSE variants of 2 sizes of the OPT arch and 1 size of the Pythia arch, including at different token scales of the babyLM benchmark. We find DYAD to be competitive (>= 90%) of DENSE performance on zero-shot (e.g. BLIMP), few-shot (OPENLM) and finetuning (GLUE) benchmarks, while being >=7-15% faster to train on-GPU even at 125m scale, besides surfacing larger speedups at increasing scale and model width.","url_abs":"https://arxiv.org/abs/2312.06881v1","url_pdf":"https://arxiv.org/pdf/2312.06881v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"dyad-a-descriptive-yet-abjuring-density","repo_url":"https://github.com/asappresearch/dyad","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"jax","reach":null}],"tasks":[{"task_slug":"descriptive","task_name":"Descriptive"},{"task_slug":null,"task_name":"GPU"}],"methods":[{"method_slug":"opt","method_name":"OPT"},{"method_slug":"pythia","method_name":"Pythia"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2312.06881","atlas_url":"https://app.syntology.ai/?focus=2312.06881","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2312.06881"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-25T09:33:49+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":[{"provenance":"deterministic:regex_extraction","url":"https://github.com/asappresearch/dyad","reach":null}],"summary":{"ran":1},"by_repo_kind":{"official":{"samples":1,"ran":1,"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":1,"samples":[{"code_sha256_prefix":"3145c5b814cc433c","entry":"Dyad","repo":"asappresearch/dyad","repo_kind":"official","path":"sparse_transformers/src/transformers/models/sparse_opt/dyad.py","file_url":"https://github.com/asappresearch/dyad/blob/HEAD/sparse_transformers/src/transformers/models/sparse_opt/dyad.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"3145c5b814cc433c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}