{"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/rwkv-x-a-linear-complexity-hybrid-language","title":"RWKV-X: A Linear Complexity Hybrid Language Model","arxiv_id":"2504.21463","date":"2025-04-30","proceeding":null,"authors":["Haowen Hou","Zhiyi Huang","Kaifeng Tan","Rongchang Lu","Fei Richard Yu"],"abstract":"In this paper, we introduce RWKV-X, a novel hybrid architecture that combines the efficiency of RWKV for short-range modeling with a sparse attention mechanism designed to capture long-range context. Unlike previous hybrid approaches that rely on full attention layers and retain quadratic complexity, RWKV-X achieves linear-time complexity in training and constant-time complexity in inference decoding. We demonstrate that RWKV-X, when continually pretrained on 64K-token sequences, achieves near-perfect accuracy on the 64K passkey retrieval benchmark. It consistently outperforms prior RWKV-7 models on long-context benchmarks, while maintaining strong performance on short-context tasks. These results highlight RWKV-X as a scalable and efficient backbone for general-purpose language modeling, capable of decoding sequences up to 1 million tokens with stable speed and memory usage. To facilitate further research and analysis, we have made the checkpoints and the associated code publicly accessible at: https://github.com/howard-hou/RWKV-X.","url_abs":"https://arxiv.org/abs/2504.21463v2","url_pdf":"https://arxiv.org/pdf/2504.21463v2.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":"rwkv-x-a-linear-complexity-hybrid-language","repo_url":"https://github.com/howard-hou/rwkv-x","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"model","task_name":"model"}],"methods":[{"method_slug":"attention","method_name":"Attention"},{"method_slug":"speed","method_name":"SPEED"},{"method_slug":"softmax","method_name":"Softmax"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":null,"atlas_url":"https://app.syntology.ai/?focus=2504.21463","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2504.21463"}},"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":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/howard-hou/rwkv-x","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":12},"by_repo_kind":{"official":{"samples":12,"ran":0,"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":"3b1aee5f3ce13a07","entry":"RWKV_x070_CMix_one","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"package/src/rwkv_x/model.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/package/src/rwkv_x/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"3b1aee5f3ce13a07"}},{"code_sha256_prefix":"449b983cd5aa924b","entry":"RWKV_x070_CMix_seq","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"package/src/rwkv_x/model.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/package/src/rwkv_x/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"449b983cd5aa924b"}},{"code_sha256_prefix":"1ce560658d1f24ec","entry":"RWKV_x070_TMix_one","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"package/src/rwkv_x/model.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/package/src/rwkv_x/model.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"1ce560658d1f24ec"}},{"code_sha256_prefix":"9f893ae6059d283e","entry":"RWKV_x070_TMix_one","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"package/src/rwkv_x/model_exp.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/package/src/rwkv_x/model_exp.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9f893ae6059d283e"}},{"code_sha256_prefix":"f0846b300e2b4057","entry":"calculate_headwise_memory_loss","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"probing/headwise_memory_loss.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/probing/headwise_memory_loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f0846b300e2b4057"}},{"code_sha256_prefix":"847084be6c83567e","entry":"compute_stats","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"probing/headwise_memory_loss.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/probing/headwise_memory_loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"847084be6c83567e"}},{"code_sha256_prefix":"e5648d7352488652","entry":"generate_prompt_landmark","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/pass_key_rwkv.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/pass_key_rwkv.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"e5648d7352488652"}},{"code_sha256_prefix":"74db7dd3f1b3abf4","entry":"load_hf","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/eval_long_loss.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/eval_long_loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"74db7dd3f1b3abf4"}},{"code_sha256_prefix":"933d4bd3c4bd2d14","entry":"load_jsonl","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/eval_long_loss.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/eval_long_loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"933d4bd3c4bd2d14"}},{"code_sha256_prefix":"60ad8220063142d0","entry":"load_rwkv","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/eval_long_loss.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/eval_long_loss.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"60ad8220063142d0"}},{"code_sha256_prefix":"5187fd0702621821","entry":"passkey_retrieval_test","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/pass_key_rwkv.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/pass_key_rwkv.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"5187fd0702621821"}},{"code_sha256_prefix":"c0d04f50608c2981","entry":"plot_heatmap","repo":"howard-hou/rwkv-x","repo_kind":"official","path":"evaluation/pass_key_rwkv.py","file_url":"https://github.com/howard-hou/rwkv-x/blob/HEAD/evaluation/pass_key_rwkv.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"c0d04f50608c2981"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}