{"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/the-pitfalls-of-next-token-prediction","title":"The pitfalls of next-token prediction","arxiv_id":"2403.06963","date":"2024-03-11","proceeding":null,"authors":["Gregor Bachmann","Vaishnavh Nagarajan"],"abstract":"Can a mere next-token predictor faithfully model human intelligence? We crystallize this emerging concern and correct popular misconceptions surrounding it, and advocate a simple multi-token objective. As a starting point, we argue that the two often-conflated phases of next-token prediction -- autoregressive inference and teacher-forced training -- must be treated distinctly. The popular criticism that errors can compound during autoregressive inference, crucially assumes that teacher-forcing has learned an accurate next-token predictor. This assumption sidesteps a more deep-rooted problem we expose: in certain classes of tasks, teacher-forcing can simply fail to learn an accurate next-token predictor in the first place. We describe a general mechanism of how teacher-forcing can fail, and design a minimal planning task where both the Transformer and the Mamba architecture empirically fail in that manner -- remarkably, despite the task being straightforward to learn. Finally, we provide preliminary evidence that this failure can be resolved using a simple modification that predicts multiple tokens in advance. We hope this finding can ground future debates and inspire explorations beyond the next-token prediction paradigm. We make our code available under https://github.com/gregorbachmann/Next-Token-Failures","url_abs":"https://arxiv.org/abs/2403.06963v2","url_pdf":"https://arxiv.org/pdf/2403.06963v2.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":"the-pitfalls-of-next-token-prediction","repo_url":"https://github.com/gregorbachmann/next-token-failures","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"mamba","task_name":"Mamba"},{"task_slug":"misconceptions","task_name":"Misconceptions"},{"task_slug":"prediction","task_name":"Prediction"}],"methods":[{"method_slug":"absolute-position-encodings","method_name":"Absolute Position Encodings"},{"method_slug":"adam","method_name":"Adam"},{"method_slug":"attention","method_name":"Attention"},{"method_slug":"bpe","method_name":"BPE"},{"method_slug":"dense-connections","method_name":"Dense Connections"},{"method_slug":"dropout","method_name":"Dropout"},{"method_slug":"label-smoothing","method_name":"Label Smoothing"},{"method_slug":"layer-normalization","method_name":"Layer Normalization"},{"method_slug":"linear-layer","method_name":"Linear Layer"},{"method_slug":"multi-head-attention","method_name":"Multi-Head Attention"},{"method_slug":"position-wise-feed-forward-layer","method_name":"Position-Wise Feed-Forward Layer"},{"method_slug":"residual-connection","method_name":"Residual Connection"},{"method_slug":"softmax","method_name":"Softmax"},{"method_slug":"transformer","method_name":"Transformer"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2403.06963","atlas_url":"https://app.syntology.ai/?focus=2403.06963","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2403.06963"}},"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":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/gregorbachmann/next-token-failures","reach":{"status":"ok"}}],"summary":{"ran":6,"unverified":1},"by_repo_kind":{"official":{"samples":7,"ran":6,"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":7,"samples":[{"code_sha256_prefix":"c1f4b077a9b230f4","entry":"accuracy","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/training_utils.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/training_utils.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":"c1f4b077a9b230f4"}},{"code_sha256_prefix":"272ef5f104a11d59","entry":"gelu","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"models/lib.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/models/lib.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":"272ef5f104a11d59"}},{"code_sha256_prefix":"1b9a2fc53c54c9be","entry":"get_lr","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/training_utils.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/training_utils.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":"1b9a2fc53c54c9be"}},{"code_sha256_prefix":"e64b971fffaa3f52","entry":"get_run_name","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/training_utils.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/training_utils.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":"e64b971fffaa3f52"}},{"code_sha256_prefix":"d8ecc47b9ccc6855","entry":"load_pythia","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/load.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/load.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":"d8ecc47b9ccc6855"}},{"code_sha256_prefix":"ca87fc1303002ce5","entry":"remap","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/load.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/load.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":"ca87fc1303002ce5"}},{"code_sha256_prefix":"33eb4da2120207ac","entry":"load_gpt","repo":"gregorbachmann/next-token-failures","repo_kind":"official","path":"utils/load.py","file_url":"https://github.com/gregorbachmann/next-token-failures/blob/HEAD/utils/load.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":"33eb4da2120207ac"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}