{"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/lattice-recurrent-unit-improving-convergence","title":"Lattice Recurrent Unit: Improving Convergence and Statistical Efficiency for Sequence Modeling","arxiv_id":"1710.02254","date":"2017-10-06","proceeding":null,"authors":["Chaitanya Ahuja","Louis-Philippe Morency"],"abstract":"Recurrent neural networks have shown remarkable success in modeling\nsequences. However low resource situations still adversely affect the\ngeneralizability of these models. We introduce a new family of models, called\nLattice Recurrent Units (LRU), to address the challenge of learning deep\nmulti-layer recurrent models with limited resources. LRU models achieve this\ngoal by creating distinct (but coupled) flow of information inside the units: a\nfirst flow along time dimension and a second flow along depth dimension. It\nalso offers a symmetry in how information can flow horizontally and vertically.\nWe analyze the effects of decoupling three different components of our LRU\nmodel: Reset Gate, Update Gate and Projected State. We evaluate this family on\nnew LRU models on computational convergence rates and statistical efficiency.\nOur experiments are performed on four publicly-available datasets, comparing\nwith Grid-LSTM and Recurrent Highway networks. Our results show that LRU has\nbetter empirical computational convergence rates and statistical efficiency\nvalues, along with learning more accurate language models.","url_abs":"http://arxiv.org/abs/1710.02254v2","url_pdf":"http://arxiv.org/pdf/1710.02254v2.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":"lattice-recurrent-unit-improving-convergence","repo_url":"https://github.com/chahuja/lru","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"lattice-recurrent-unit-improving-convergence","repo_url":"https://github.com/MindCode-4/code-7/tree/main/lru","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null},{"paper_slug":"lattice-recurrent-unit-improving-convergence","repo_url":"https://github.com/MindSpore-scientific/code-6/tree/main/lru","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":0,"framework":"mindspore","reach":null}],"tasks":[],"methods":[{"method_slug":"highway-networks","method_name":"Highway networks"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":null,"mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}