{"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/dive-into-the-chasm-probing-the-gap-between","title":"Dive into the Chasm: Probing the Gap between In- and Cross-Topic Generalization","arxiv_id":"2402.01375","date":"2024-02-02","proceeding":null,"authors":["Andreas Waldis","Yufang Hou","Iryna Gurevych"],"abstract":"Pre-trained language models (LMs) perform well in In-Topic setups, where training and testing data come from the same topics. However, they face challenges in Cross-Topic scenarios where testing data is derived from distinct topics -- such as Gun Control. This study analyzes various LMs with three probing-based experiments to shed light on the reasons behind the In- vs. Cross-Topic generalization gap. Thereby, we demonstrate, for the first time, that generalization gaps and the robustness of the embedding space vary significantly across LMs. Additionally, we assess larger LMs and underscore the relevance of our analysis for recent models. Overall, diverse pre-training objectives, architectural regularization, or data deduplication contribute to more robust LMs and diminish generalization gaps. Our research contributes to a deeper understanding and comparison of language models across different generalization scenarios.","url_abs":"https://arxiv.org/abs/2402.01375v1","url_pdf":"https://arxiv.org/pdf/2402.01375v1.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":"dive-into-the-chasm-probing-the-gap-between","repo_url":"https://github.com/ukplab/eacl2024-cross-topic-probing","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2402.01375","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.01375"}},"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/ukplab/eacl2024-cross-topic-probing","reach":{"status":"ok","spdx":"Apache-2.0"}}],"summary":{"ran":9,"unverified":3},"by_repo_kind":{"official":{"samples":12,"ran":9,"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":"0dc7954105dd5021","entry":"assign_labels","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/control.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/control.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"0dc7954105dd5021"}},{"code_sha256_prefix":"4e602ab97e73567f","entry":"forward_from_specific_layer","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/bert_encoding.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/bert_encoding.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"4e602ab97e73567f"}},{"code_sha256_prefix":"4c42d81f47c0a093","entry":"get_num_labels","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/control.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/control.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"4c42d81f47c0a093"}},{"code_sha256_prefix":"f51466d4f47f3d3d","entry":"get_projection_to_intersection_of_nullspaces","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/debias/debias.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/debias/debias.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"f51466d4f47f3d3d"}},{"code_sha256_prefix":"8df841b6db2c9c8a","entry":"get_rowspace_projection","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/debias/debias.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/debias/debias.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"8df841b6db2c9c8a"}},{"code_sha256_prefix":"1e2440ad06038e17","entry":"label_sentences","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/control.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/control.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"1e2440ad06038e17"}},{"code_sha256_prefix":"76f97aa1b9b381f6","entry":"load_config","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/parsing/convert_cross_topic_tasks.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/parsing/convert_cross_topic_tasks.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"76f97aa1b9b381f6"}},{"code_sha256_prefix":"e57d2979287b15b3","entry":"load_data","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/debiased_finetuning/utils.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/debiased_finetuning/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"e57d2979287b15b3"}},{"code_sha256_prefix":"6686bf211e128ab1","entry":"load_labels","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/debiased_finetuning/utils.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/debiased_finetuning/utils.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6686bf211e128ab1"}},{"code_sha256_prefix":"32cf84be209e2a2e","entry":"debias_by_specific_directions","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/debias/debias.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/debias/debias.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"32cf84be209e2a2e"}},{"code_sha256_prefix":"6d612990f1913f51","entry":"lm_encoding","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/bert_encoding.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/bert_encoding.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"6d612990f1913f51"}},{"code_sha256_prefix":"35bf6ca8949889cf","entry":"lm_masked_encoding","repo":"ukplab/eacl2024-cross-topic-probing","repo_kind":"official","path":"src/amnesic_probing/encoders/bert_encoding.py","file_url":"https://github.com/ukplab/eacl2024-cross-topic-probing/blob/HEAD/src/amnesic_probing/encoders/bert_encoding.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"Apache-2.0","inline_ok":true,"mcp_get_code":{"code_sha256":"35bf6ca8949889cf"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}