{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:QL2QQK3U4L6RDMGSL7MVSV4JSP","short_pith_number":"pith:QL2QQK3U","schema_version":"1.0","canonical_sha256":"82f5082b74e2fd11b0d25fd959578993d56e4594565f5a23ef39c92e07cbb394","source":{"kind":"arxiv","id":"2310.18862","version":1},"attestation_state":"computed","paper":{"title":"Counterfactually Probing Language Identity in Multilingual Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Anirudh Srinivasan, Kyle Mahowald, Venkata S Govindarajan","submitted_at":"2023-10-29T01:21:36Z","abstract_excerpt":"Techniques in causal analysis of language models illuminate how linguistic information is organized in LLMs. We use one such technique, AlterRep, a method of counterfactual probing, to explore the internal structure of multilingual models (mBERT and XLM-R). We train a linear classifier on a binary language identity task, to classify tokens between Language X and Language Y. Applying a counterfactual probing procedure, we use the classifier weights to project the embeddings into the null space and push the resulting embeddings either in the direction of Language X or Language Y. Then we evaluat"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2310.18862","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-29T01:21:36Z","cross_cats_sorted":[],"title_canon_sha256":"77aa420d0fc145dc01702ea588f97c5a8847f61c854363858a72352b3ad1755b","abstract_canon_sha256":"2a6d78fa80ad67f2d1f9270c9914014d4cb36f51d110b5ef91008c4de1aab91b"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:06:42.563466Z","signature_b64":"J/r8iFYbNXTYcK8O8NHbOoIWkfKm+8lI2hU/XL1pjtWMJtjy4epsDd3WGkJcwsfTJaysEoezVEETDd2UbSVqBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"82f5082b74e2fd11b0d25fd959578993d56e4594565f5a23ef39c92e07cbb394","last_reissued_at":"2026-07-05T07:06:42.563055Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:06:42.563055Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Counterfactually Probing Language Identity in Multilingual Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Anirudh Srinivasan, Kyle Mahowald, Venkata S Govindarajan","submitted_at":"2023-10-29T01:21:36Z","abstract_excerpt":"Techniques in causal analysis of language models illuminate how linguistic information is organized in LLMs. We use one such technique, AlterRep, a method of counterfactual probing, to explore the internal structure of multilingual models (mBERT and XLM-R). We train a linear classifier on a binary language identity task, to classify tokens between Language X and Language Y. Applying a counterfactual probing procedure, we use the classifier weights to project the embeddings into the null space and push the resulting embeddings either in the direction of Language X or Language Y. Then we evaluat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.18862","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2310.18862/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2310.18862","created_at":"2026-07-05T07:06:42.563113+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.18862v1","created_at":"2026-07-05T07:06:42.563113+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.18862","created_at":"2026-07-05T07:06:42.563113+00:00"},{"alias_kind":"pith_short_12","alias_value":"QL2QQK3U4L6R","created_at":"2026-07-05T07:06:42.563113+00:00"},{"alias_kind":"pith_short_16","alias_value":"QL2QQK3U4L6RDMGS","created_at":"2026-07-05T07:06:42.563113+00:00"},{"alias_kind":"pith_short_8","alias_value":"QL2QQK3U","created_at":"2026-07-05T07:06:42.563113+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP","json":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP.json","graph_json":"https://pith.science/api/pith-number/QL2QQK3U4L6RDMGSL7MVSV4JSP/graph.json","events_json":"https://pith.science/api/pith-number/QL2QQK3U4L6RDMGSL7MVSV4JSP/events.json","paper":"https://pith.science/paper/QL2QQK3U"},"agent_actions":{"view_html":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP","download_json":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP.json","view_paper":"https://pith.science/paper/QL2QQK3U","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.18862&json=true","fetch_graph":"https://pith.science/api/pith-number/QL2QQK3U4L6RDMGSL7MVSV4JSP/graph.json","fetch_events":"https://pith.science/api/pith-number/QL2QQK3U4L6RDMGSL7MVSV4JSP/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP/action/timestamp_anchor","attest_storage":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP/action/storage_attestation","attest_author":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP/action/author_attestation","sign_citation":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP/action/citation_signature","submit_replication":"https://pith.science/pith/QL2QQK3U4L6RDMGSL7MVSV4JSP/action/replication_record"}},"created_at":"2026-07-05T07:06:42.563113+00:00","updated_at":"2026-07-05T07:06:42.563113+00:00"}