{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:IBDV3TJ256PTAZZZV3QJNZUKNF","short_pith_number":"pith:IBDV3TJ2","schema_version":"1.0","canonical_sha256":"40475dcd3aef9f306739aee096e68a6967c35955127287a171e296fa63adce08","source":{"kind":"arxiv","id":"2506.01074","version":1},"attestation_state":"computed","paper":{"title":"How Programming Concepts and Neurons Are Shared in Code Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.PL","cs.SE"],"primary_cat":"cs.CL","authors_text":"Amir Hossein Kargaran, Fran\\c{c}ois Yvon, Hinrich Sch\\\"utze, Yihong Liu","submitted_at":"2025-06-01T16:24:13Z","abstract_excerpt":"Several studies have explored the mechanisms of large language models (LLMs) in coding tasks, but most have focused on programming languages (PLs) in a monolingual setting. In this paper, we investigate the relationship between multiple PLs and English in the concept space of LLMs. We perform a few-shot translation task on 21 PL pairs using two Llama-based models. By decoding the embeddings of intermediate layers during this task, we observe that the concept space is closer to English (including PL keywords) and assigns high probabilities to English tokens in the second half of the intermediat"},"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":"2506.01074","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2025-06-01T16:24:13Z","cross_cats_sorted":["cs.PL","cs.SE"],"title_canon_sha256":"42c1b480e151277350aa354c2791aab6e9b9a0d4eba9825b8ab6204ca6f920a4","abstract_canon_sha256":"bdebeb714c2a51c1ce9347c0d53ac2845f1b02c03c31156daf7bcc964599f962"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:13:45.620099Z","signature_b64":"zpukpdmBRCnTuC5SyyvFoDp3rtPE6HFf/2smxvtRh72VCvApLlI49zPgeRjP784aYNCtwMdOOXGd0HZIRM+6AA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"40475dcd3aef9f306739aee096e68a6967c35955127287a171e296fa63adce08","last_reissued_at":"2026-07-05T11:13:45.619628Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:13:45.619628Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"How Programming Concepts and Neurons Are Shared in Code Language Models","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.PL","cs.SE"],"primary_cat":"cs.CL","authors_text":"Amir Hossein Kargaran, Fran\\c{c}ois Yvon, Hinrich Sch\\\"utze, Yihong Liu","submitted_at":"2025-06-01T16:24:13Z","abstract_excerpt":"Several studies have explored the mechanisms of large language models (LLMs) in coding tasks, but most have focused on programming languages (PLs) in a monolingual setting. In this paper, we investigate the relationship between multiple PLs and English in the concept space of LLMs. We perform a few-shot translation task on 21 PL pairs using two Llama-based models. By decoding the embeddings of intermediate layers during this task, we observe that the concept space is closer to English (including PL keywords) and assigns high probabilities to English tokens in the second half of the intermediat"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.01074","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/2506.01074/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":"2506.01074","created_at":"2026-07-05T11:13:45.619687+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.01074v1","created_at":"2026-07-05T11:13:45.619687+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.01074","created_at":"2026-07-05T11:13:45.619687+00:00"},{"alias_kind":"pith_short_12","alias_value":"IBDV3TJ256PT","created_at":"2026-07-05T11:13:45.619687+00:00"},{"alias_kind":"pith_short_16","alias_value":"IBDV3TJ256PTAZZZ","created_at":"2026-07-05T11:13:45.619687+00:00"},{"alias_kind":"pith_short_8","alias_value":"IBDV3TJ2","created_at":"2026-07-05T11:13:45.619687+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2601.14004","citing_title":"Locate, Steer, and Improve: A Practical Survey of Actionable Mechanistic Interpretability in Large Language Models","ref_index":146,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF","json":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF.json","graph_json":"https://pith.science/api/pith-number/IBDV3TJ256PTAZZZV3QJNZUKNF/graph.json","events_json":"https://pith.science/api/pith-number/IBDV3TJ256PTAZZZV3QJNZUKNF/events.json","paper":"https://pith.science/paper/IBDV3TJ2"},"agent_actions":{"view_html":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF","download_json":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF.json","view_paper":"https://pith.science/paper/IBDV3TJ2","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.01074&json=true","fetch_graph":"https://pith.science/api/pith-number/IBDV3TJ256PTAZZZV3QJNZUKNF/graph.json","fetch_events":"https://pith.science/api/pith-number/IBDV3TJ256PTAZZZV3QJNZUKNF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF/action/storage_attestation","attest_author":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF/action/author_attestation","sign_citation":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF/action/citation_signature","submit_replication":"https://pith.science/pith/IBDV3TJ256PTAZZZV3QJNZUKNF/action/replication_record"}},"created_at":"2026-07-05T11:13:45.619687+00:00","updated_at":"2026-07-05T11:13:45.619687+00:00"}