{"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/information-theoretic-probing-with-minimum","title":"Information-Theoretic Probing with Minimum Description Length","arxiv_id":"2003.12298","date":"2020-03-27","proceeding":"EMNLP 2020 11","authors":["Elena Voita","Ivan Titov"],"abstract":"To measure how well pretrained representations encode some linguistic property, it is common to use accuracy of a probe, i.e. a classifier trained to predict the property from the representations. Despite widespread adoption of probes, differences in their accuracy fail to adequately reflect differences in representations. For example, they do not substantially favour pretrained representations over randomly initialized ones. Analogously, their accuracy can be similar when probing for genuine linguistic labels and probing for random synthetic tasks. To see reasonable differences in accuracy with respect to these random baselines, previous work had to constrain either the amount of probe training data or its model size. Instead, we propose an alternative to the standard probes, information-theoretic probing with minimum description length (MDL). With MDL probing, training a probe to predict labels is recast as teaching it to effectively transmit the data. Therefore, the measure of interest changes from probe accuracy to the description length of labels given representations. In addition to probe quality, the description length evaluates \"the amount of effort\" needed to achieve the quality. This amount of effort characterizes either (i) size of a probing model, or (ii) the amount of data needed to achieve the high quality. We consider two methods for estimating MDL which can be easily implemented on top of the standard probing pipelines: variational coding and online coding. We show that these methods agree in results and are more informative and stable than the standard probes.","url_abs":"https://arxiv.org/abs/2003.12298v1","url_pdf":"https://arxiv.org/pdf/2003.12298v1.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":"information-theoretic-probing-with-minimum","repo_url":"https://github.com/lena-voita/description-length-probing","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"information-theoretic-probing-with-minimum","repo_url":"https://github.com/willwhitney/reprieve","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[],"methods":[{"method_slug":"mdl","method_name":"MDL"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2003.12298","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2003.12298"}},"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/lena-voita/description-length-probing","reach":null},{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/willwhitney/reprieve","reach":null}],"summary":{"ran":1,"ran_fixture":1,"unverified":3},"by_repo_kind":{"official":{"samples":4,"ran":2,"repositories":1},"listed":{"samples":1,"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":5,"samples":[{"code_sha256_prefix":"f5f58450c18992a1","entry":"Probe","repo":"lena-voita/description-length-probing","repo_kind":"official","path":"control_tasks/control_tasks/probe.py","file_url":"https://github.com/lena-voita/description-length-probing/blob/HEAD/control_tasks/control_tasks/probe.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":"f5f58450c18992a1"}},{"code_sha256_prefix":"2466d861eece0826","entry":"reparametrize","repo":"lena-voita/description-length-probing","repo_kind":"official","path":"control_tasks/control_tasks/probe.py","file_url":"https://github.com/lena-voita/description-length-probing/blob/HEAD/control_tasks/control_tasks/probe.py","link_basis":"first_harvest_node","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"2466d861eece0826"}},{"code_sha256_prefix":"850fa1698390a72a","entry":"LinearGroupNJ","repo":"lena-voita/description-length-probing","repo_kind":"official","path":"control_tasks/control_tasks/probe.py","file_url":"https://github.com/lena-voita/description-length-probing/blob/HEAD/control_tasks/control_tasks/probe.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":"850fa1698390a72a"}},{"code_sha256_prefix":"0e3860a096fe4f95","entry":"LossDataEstimator","repo":"willwhitney/reprieve","repo_kind":"listed","path":"reprieve/api.py","file_url":"https://github.com/willwhitney/reprieve/blob/HEAD/reprieve/api.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":"0e3860a096fe4f95"}},{"code_sha256_prefix":"2d967d1afb007373","entry":"OneWordNNLabelProbeBayesCompression","repo":"lena-voita/description-length-probing","repo_kind":"official","path":"control_tasks/control_tasks/probe.py","file_url":"https://github.com/lena-voita/description-length-probing/blob/HEAD/control_tasks/control_tasks/probe.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":"2d967d1afb007373"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}