{"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/faith-and-fate-limits-of-transformers-on-1","title":"Faith and Fate: Limits of Transformers on Compositionality","arxiv_id":"2305.18654","date":"2023-05-29","proceeding":"NeurIPS 2023 11","authors":["Nouha Dziri","Ximing Lu","Melanie Sclar","Xiang Lorraine Li","Liwei Jiang","Bill Yuchen Lin","Peter West","Chandra Bhagavatula","Ronan Le Bras","Jena D. Hwang","Soumya Sanyal","Sean Welleck","Xiang Ren","Allyson Ettinger","Zaid Harchaoui","Yejin Choi"],"abstract":"Transformer large language models (LLMs) have sparked admiration for their exceptional performance on tasks that demand intricate multi-step reasoning. Yet, these models simultaneously show failures on surprisingly trivial problems. This begs the question: Are these errors incidental, or do they signal more substantial limitations? In an attempt to demystify transformer LLMs, we investigate the limits of these models across three representative compositional tasks -- multi-digit multiplication, logic grid puzzles, and a classic dynamic programming problem. These tasks require breaking problems down into sub-steps and synthesizing these steps into a precise answer. We formulate compositional tasks as computation graphs to systematically quantify the level of complexity, and break down reasoning steps into intermediate sub-procedures. Our empirical findings suggest that transformer LLMs solve compositional tasks by reducing multi-step compositional reasoning into linearized subgraph matching, without necessarily developing systematic problem-solving skills. To round off our empirical study, we provide theoretical arguments on abstract multi-step reasoning problems that highlight how autoregressive generations' performance can rapidly decay with\\,increased\\,task\\,complexity.","url_abs":"https://arxiv.org/abs/2305.18654v3","url_pdf":"https://arxiv.org/pdf/2305.18654v3.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":"faith-and-fate-limits-of-transformers-on-1","repo_url":"https://github.com/nouhadziri/faith-and-fate","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"none","reach":null}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/2305.18654","atlas_url":"https://app.syntology.ai/?focus=2305.18654","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2305.18654"}},"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/nouhadziri/faith-and-fate","reach":null}],"summary":{"ran_honours":3,"ran_fixture":1,"ran_draft_wrong":1},"by_repo_kind":{"official":{"samples":5,"ran":5,"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":"520b18e6f4c86711","entry":"all_n_digit","repo":"nouhadziri/faith-and-fate","repo_kind":"official","path":"multiplication/build_data.py","file_url":"https://github.com/nouhadziri/faith-and-fate/blob/HEAD/multiplication/build_data.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"520b18e6f4c86711"}},{"code_sha256_prefix":"754e258c84562c66","entry":"cartesian","repo":"nouhadziri/faith-and-fate","repo_kind":"official","path":"multiplication/build_data.py","file_url":"https://github.com/nouhadziri/faith-and-fate/blob/HEAD/multiplication/build_data.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"754e258c84562c66"}},{"code_sha256_prefix":"f300cecd9a71e8ef","entry":"generate_scratchpad_completion","repo":"nouhadziri/faith-and-fate","repo_kind":"official","path":"dynamic_programming/generate_training_data.py","file_url":"https://github.com/nouhadziri/faith-and-fate/blob/HEAD/dynamic_programming/generate_training_data.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"deterministic","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"f300cecd9a71e8ef"}},{"code_sha256_prefix":"9d9f3f1af5a83331","entry":"get_idx2clue","repo":"nouhadziri/faith-and-fate","repo_kind":"official","path":"logic_puzzle/graph/reasoning_path.py","file_url":"https://github.com/nouhadziri/faith-and-fate/blob/HEAD/logic_puzzle/graph/reasoning_path.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"OUTPUT_MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"9d9f3f1af5a83331"}},{"code_sha256_prefix":"62e9033feec5f17e","entry":"random_n_digit","repo":"nouhadziri/faith-and-fate","repo_kind":"official","path":"multiplication/build_data.py","file_url":"https://github.com/nouhadziri/faith-and-fate/blob/HEAD/multiplication/build_data.py","link_basis":"first_harvest_node","language":"python","status":"ran_honours","verification_level":1,"contract_check":"HONOURS","metamorphic_tier":"well_formed","behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"62e9033feec5f17e"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}