{"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/avoiding-reasoning-shortcuts-adversarial","title":"Avoiding Reasoning Shortcuts: Adversarial Evaluation, Training, and Model Development for Multi-Hop QA","arxiv_id":"1906.07132","date":"2019-06-17","proceeding":"ACL 2019 7","authors":["Yichen Jiang","Mohit Bansal"],"abstract":"Multi-hop question answering requires a model to connect multiple pieces of evidence scattered in a long context to answer the question. In this paper, we show that in the multi-hop HotpotQA (Yang et al., 2018) dataset, the examples often contain reasoning shortcuts through which models can directly locate the answer by word-matching the question with a sentence in the context. We demonstrate this issue by constructing adversarial documents that create contradicting answers to the shortcut but do not affect the validity of the original answer. The performance of strong baseline models drops significantly on our adversarial evaluation, indicating that they are indeed exploiting the shortcuts rather than performing multi-hop reasoning. After adversarial training, the baseline's performance improves but is still limited on the adversarial evaluation. Hence, we use a control unit that dynamically attends to the question at different reasoning hops to guide the model's multi-hop reasoning. We show that this 2-hop model trained on the regular data is more robust to the adversaries than the baseline model. After adversarial training, this 2-hop model not only achieves improvements over its counterpart trained on regular data, but also outperforms the adversarially-trained 1-hop baseline. We hope that these insights and initial improvements will motivate the development of new models that combine explicit compositional reasoning with adversarial training.","url_abs":"https://arxiv.org/abs/1906.07132v1","url_pdf":"https://arxiv.org/pdf/1906.07132v1.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":"avoiding-reasoning-shortcuts-adversarial","repo_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"multi-hop-question-answering","task_name":"Multi-hop Question Answering"},{"task_slug":"question-answering","task_name":"Question Answering"},{"task_slug":"sentence","task_name":"Sentence"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1906.07132","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"1906.07132"}},"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/jiangycTarheel/Adversarial-MultiHopQA","reach":{"status":"ok","spdx":"MIT"}}],"summary":{"unverified":13},"by_repo_kind":{"official":{"samples":13,"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":"52f030918856dc44","entry":"aggregate_metrics","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nectar/nectar/theanoutil/model.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nectar/nectar/theanoutil/model.py","link_basis":"first_harvest_node","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":"52f030918856dc44"}},{"code_sha256_prefix":"b531959d4a3b2b6b","entry":"build_word_vector_matrix","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/utils.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/utils.py","link_basis":"first_harvest_node","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":"b531959d4a3b2b6b"}},{"code_sha256_prefix":"2e226d735609d4e1","entry":"find_nearest","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/get_nearest.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/get_nearest.py","link_basis":"first_harvest_node","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":"2e226d735609d4e1"}},{"code_sha256_prefix":"0d4400d162743098","entry":"format_epoch_str","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nectar/nectar/theanoutil/model.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nectar/nectar/theanoutil/model.py","link_basis":"first_harvest_node","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":"0d4400d162743098"}},{"code_sha256_prefix":"47eea3afc458a05e","entry":"get_cache_filename_from_args","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/utils.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/utils.py","link_basis":"first_harvest_node","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":"47eea3afc458a05e"}},{"code_sha256_prefix":"c7eebf1fb92eae8e","entry":"get_label_dictionaries","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/utils.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/utils.py","link_basis":"first_harvest_node","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":"c7eebf1fb92eae8e"}},{"code_sha256_prefix":"f609cb7c4f3715ca","entry":"load_ans_title_cache","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"convert_sp_facts.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/convert_sp_facts.py","link_basis":"first_harvest_node","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":"f609cb7c4f3715ca"}},{"code_sha256_prefix":"4bc6c631e9149f9d","entry":"load_cache","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"convert_sp_facts.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/convert_sp_facts.py","link_basis":"first_harvest_node","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":"4bc6c631e9149f9d"}},{"code_sha256_prefix":"5dd05013737324f6","entry":"longest_common_substring","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/get_nearest.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/get_nearest.py","link_basis":"first_harvest_node","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":"5dd05013737324f6"}},{"code_sha256_prefix":"de3171c2eb7de98f","entry":"lookup_answer_generate","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"answer_rules.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/answer_rules.py","link_basis":"first_harvest_node","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":"de3171c2eb7de98f"}},{"code_sha256_prefix":"7e97f29471939565","entry":"lookup_title_generate","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"bridge_entity_rules.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/bridge_entity_rules.py","link_basis":"first_harvest_node","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":"7e97f29471939565"}},{"code_sha256_prefix":"d76d37fbb1793177","entry":"rejoin","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nectar/nectar/corenlp/util.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nectar/nectar/corenlp/util.py","link_basis":"first_harvest_node","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":"d76d37fbb1793177"}},{"code_sha256_prefix":"5eb2a8cac799f452","entry":"word_arithmetic","repo":"jiangycTarheel/Adversarial-MultiHopQA","repo_kind":"official","path":"nearest_glove/get_nearest.py","file_url":"https://github.com/jiangycTarheel/Adversarial-MultiHopQA/blob/HEAD/nearest_glove/get_nearest.py","link_basis":"first_harvest_node","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":"5eb2a8cac799f452"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}