{"id":"2411.12992","title":"MemoryFormer: Minimize Transformer Computation by Removing Fully-Connected Layers","url":"https://codeofpaper.com/paper/2411.12992","arxiv_url":"https://arxiv.org/abs/2411.12992","published":"2024-11-01","code":"community","best":{"full_name":"C0-Design/MemoryFormer","url":"https://github.com/C0-Design/MemoryFormer","tier":"high_confidence_community","stars":16,"license":null,"framework":null,"language":"Python","last_commit":"2026-03-24T15:42:29"},"most_starred_community":null,"implementations":[{"full_name":"C0-Design/MemoryFormer","url":"https://github.com/C0-Design/MemoryFormer","tier":"high_confidence_community","stars":16,"license":null,"framework":null,"language":"Python","last_commit":"2026-03-24T15:42:29"}],"possible_matches":0}