{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NW2USII44HY7EG3RYG3537JMZT","short_pith_number":"pith:NW2USII4","schema_version":"1.0","canonical_sha256":"6db549211ce1f1f21b71c1b7ddfd2cccfe4d9da754af18909120195e6036a83b","source":{"kind":"arxiv","id":"2404.09937","version":2},"attestation_state":"computed","paper":{"title":"Compression Represents Intelligence Linearly","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IT","cs.LG","math.IT"],"primary_cat":"cs.CL","authors_text":"Jinghan Zhang, Junxian He, Yuzhen Huang, Zifei Shan","submitted_at":"2024-04-15T17:03:41Z","abstract_excerpt":"There is a belief that learning to compress well will lead to intelligence. Recently, language modeling has been shown to be equivalent to compression, which offers a compelling rationale for the success of large language models (LLMs): the development of more advanced language models is essentially enhancing compression which facilitates intelligence. Despite such appealing discussions, little empirical evidence is present for the interplay between compression and intelligence. In this work, we examine their relationship in the context of LLMs, treating LLMs as data compressors. Given the abs"},"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":"2404.09937","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-04-15T17:03:41Z","cross_cats_sorted":["cs.AI","cs.IT","cs.LG","math.IT"],"title_canon_sha256":"08448d8fe2c2f528753055c5707d75589e0086bb4d48d5f865e08dd56e7a2b9c","abstract_canon_sha256":"14ce3222285f17b94e8ca5c239d36ca5b307f69ef5cf843b6070595fcd058517"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:56:47.286433Z","signature_b64":"NzKQxywTFLOHEXlH/E4fEBpYihmB/a39RX/1XcweA4tLs88+hPJV4m9nnDTqlT6OJg6CTL7MK/mPQoCe4T3HCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6db549211ce1f1f21b71c1b7ddfd2cccfe4d9da754af18909120195e6036a83b","last_reissued_at":"2026-07-05T08:56:47.285875Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:56:47.285875Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Compression Represents Intelligence Linearly","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.IT","cs.LG","math.IT"],"primary_cat":"cs.CL","authors_text":"Jinghan Zhang, Junxian He, Yuzhen Huang, Zifei Shan","submitted_at":"2024-04-15T17:03:41Z","abstract_excerpt":"There is a belief that learning to compress well will lead to intelligence. Recently, language modeling has been shown to be equivalent to compression, which offers a compelling rationale for the success of large language models (LLMs): the development of more advanced language models is essentially enhancing compression which facilitates intelligence. Despite such appealing discussions, little empirical evidence is present for the interplay between compression and intelligence. In this work, we examine their relationship in the context of LLMs, treating LLMs as data compressors. Given the abs"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.09937","kind":"arxiv","version":2},"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/2404.09937/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":"2404.09937","created_at":"2026-07-05T08:56:47.285939+00:00"},{"alias_kind":"arxiv_version","alias_value":"2404.09937v2","created_at":"2026-07-05T08:56:47.285939+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.09937","created_at":"2026-07-05T08:56:47.285939+00:00"},{"alias_kind":"pith_short_12","alias_value":"NW2USII44HY7","created_at":"2026-07-05T08:56:47.285939+00:00"},{"alias_kind":"pith_short_16","alias_value":"NW2USII44HY7EG3R","created_at":"2026-07-05T08:56:47.285939+00:00"},{"alias_kind":"pith_short_8","alias_value":"NW2USII4","created_at":"2026-07-05T08:56:47.285939+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.15079","citing_title":"Ling and Ring 2.6 Technical Report: Efficient and Instant Agentic Intelligence at Trillion-Parameter Scale","ref_index":69,"is_internal_anchor":false},{"citing_arxiv_id":"2606.06273","citing_title":"Adapting Diffusion Language Models for Lossless Pixel-Level Image Transmission","ref_index":6,"is_internal_anchor":false},{"citing_arxiv_id":"2601.20255","citing_title":"HE-SNR: Uncovering Latent Logic via Entropy for Guiding Mid-Training on SWE-bench","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09724","citing_title":"Model Capacity Determines Grokking through Competing Memorisation and Generalisation Speeds","ref_index":9,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06548","citing_title":"Continuous Latent Diffusion Language Model","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2502.09992","citing_title":"Large Language Diffusion Models","ref_index":14,"is_internal_anchor":false},{"citing_arxiv_id":"2604.20940","citing_title":"Sema: Semantic Transport for Real-Time Multimodal Agents","ref_index":19,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT","json":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT.json","graph_json":"https://pith.science/api/pith-number/NW2USII44HY7EG3RYG3537JMZT/graph.json","events_json":"https://pith.science/api/pith-number/NW2USII44HY7EG3RYG3537JMZT/events.json","paper":"https://pith.science/paper/NW2USII4"},"agent_actions":{"view_html":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT","download_json":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT.json","view_paper":"https://pith.science/paper/NW2USII4","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2404.09937&json=true","fetch_graph":"https://pith.science/api/pith-number/NW2USII44HY7EG3RYG3537JMZT/graph.json","fetch_events":"https://pith.science/api/pith-number/NW2USII44HY7EG3RYG3537JMZT/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT/action/storage_attestation","attest_author":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT/action/author_attestation","sign_citation":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT/action/citation_signature","submit_replication":"https://pith.science/pith/NW2USII44HY7EG3RYG3537JMZT/action/replication_record"}},"created_at":"2026-07-05T08:56:47.285939+00:00","updated_at":"2026-07-05T08:56:47.285939+00:00"}