{"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/do-language-models-have-beliefs-methods-for","title":"Do Language Models Have Beliefs? Methods for Detecting, Updating, and Visualizing Model Beliefs","arxiv_id":"2111.13654","date":"2021-11-26","proceeding":null,"authors":["Peter Hase","Mona Diab","Asli Celikyilmaz","Xian Li","Zornitsa Kozareva","Veselin Stoyanov","Mohit Bansal","Srinivasan Iyer"],"abstract":"Do language models have beliefs about the world? Dennett (1995) famously argues that even thermostats have beliefs, on the view that a belief is simply an informational state decoupled from any motivational state. In this paper, we discuss approaches to detecting when models have beliefs about the world, and we improve on methods for updating model beliefs to be more truthful, with a focus on methods based on learned optimizers or hypernetworks. Our main contributions include: (1) new metrics for evaluating belief-updating methods that focus on the logical consistency of beliefs, (2) a training objective for Sequential, Local, and Generalizing model updates (SLAG) that improves the performance of learned optimizers, and (3) the introduction of the belief graph, which is a new form of interface with language models that shows the interdependencies between model beliefs. Our experiments suggest that models possess belief-like qualities to only a limited extent, but update methods can both fix incorrect model beliefs and greatly improve their consistency. Although off-the-shelf optimizers are surprisingly strong belief-updating baselines, our learned optimizers can outperform them in more difficult settings than have been considered in past work. Code is available at https://github.com/peterbhase/SLAG-Belief-Updating","url_abs":"https://arxiv.org/abs/2111.13654v1","url_pdf":"https://arxiv.org/pdf/2111.13654v1.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":"do-language-models-have-beliefs-methods-for","repo_url":"https://github.com/peterbhase/slag-belief-updating","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2111.13654","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2111.13654"}},"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/peterbhase/slag-belief-updating","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"ran":1,"unverified":11},"by_repo_kind":{"official":{"samples":12,"ran":1,"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":"ecdceb0edcf86391","entry":"p_value","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"statistical_analysis.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/statistical_analysis.py","link_basis":"harvester_set","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"ecdceb0edcf86391"}},{"code_sha256_prefix":"4bd0b9f5f12663b0","entry":"add_line_breaks_to_sentence","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"graph_utils.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/graph_utils.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":"4bd0b9f5f12663b0"}},{"code_sha256_prefix":"1068a66c77c7604f","entry":"command_to_dict","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"utils.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/utils.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":"1068a66c77c7604f"}},{"code_sha256_prefix":"a6ed60e4ae0bead5","entry":"element_is_primitive","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"statistical_analysis.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/statistical_analysis.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":"a6ed60e4ae0bead5"}},{"code_sha256_prefix":"60159b5b4d19fa95","entry":"filter_for_rare_relations","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"data_utils/filter_wikidata.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/data_utils/filter_wikidata.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":"60159b5b4d19fa95"}},{"code_sha256_prefix":"ad365ef21c622fc9","entry":"get_experiment_name_from_command","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"utils.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/utils.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":"ad365ef21c622fc9"}},{"code_sha256_prefix":"f555c3b4b4784251","entry":"get_result_sheet_name","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"utils.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/utils.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":"f555c3b4b4784251"}},{"code_sha256_prefix":"f4cecdbe8b090444","entry":"jaccard_similarity","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"metrics.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/metrics.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":"f4cecdbe8b090444"}},{"code_sha256_prefix":"bd9fe803bb4b72c0","entry":"nx_graph_from_pd_df","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"graph_utils.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/graph_utils.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":"bd9fe803bb4b72c0"}},{"code_sha256_prefix":"c32afabea331f1dd","entry":"pull_independent_acc_when_wrong","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"metrics.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/metrics.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":"c32afabea331f1dd"}},{"code_sha256_prefix":"531531be2eb535af","entry":"to_pseudo_language","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"data_utils/shuffle_leapofthought_splits.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/data_utils/shuffle_leapofthought_splits.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":"531531be2eb535af"}},{"code_sha256_prefix":"75bc7f3f6f994a7c","entry":"unroll_stats","repo":"peterbhase/slag-belief-updating","repo_kind":"official","path":"statistical_analysis.py","file_url":"https://github.com/peterbhase/slag-belief-updating/blob/HEAD/statistical_analysis.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":"75bc7f3f6f994a7c"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}