{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:NTGWAWUKTPFQDJAM4WWGLQFIUY","short_pith_number":"pith:NTGWAWUK","schema_version":"1.0","canonical_sha256":"6ccd605a8a9bcb01a40ce5ac65c0a8a61bedb70cc1d9ab141241dce075c14aa9","source":{"kind":"arxiv","id":"2405.12617","version":2},"attestation_state":"computed","paper":{"title":"Quantifying Semantic Emergence in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hang Chen, Jiaying Zhu, Wenya Wang, Xinyu Yang","submitted_at":"2024-05-21T09:12:20Z","abstract_excerpt":"Large language models (LLMs) are widely recognized for their exceptional capacity to capture semantics meaning. Yet, there remains no established metric to quantify this capability. In this work, we introduce a quantitative metric, Information Emergence (IE), designed to measure LLMs' ability to extract semantics from input tokens. We formalize ``semantics'' as the meaningful information abstracted from a sequence of tokens and quantify this by comparing the entropy reduction observed for a sequence of tokens (macro-level) and individual tokens (micro-level). To achieve this, we design a light"},"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":"2405.12617","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-05-21T09:12:20Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"f2652a3b2e4927057d96fc61e8f09276f6c6add5bb1d5e08018506beecc0a6f7","abstract_canon_sha256":"856b412c1504acca0d7140841cff81de710a0f5959a896dc787d92f3026b6314"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:50:47.529633Z","signature_b64":"RBE1rg3VOOAvK6ot6PAJDmETjc1e2W1a29zYybmHn+TxL3vuHDNg37OrfTwqqUX9PknwRtFvLWs4VfRvRDVBDw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"6ccd605a8a9bcb01a40ce5ac65c0a8a61bedb70cc1d9ab141241dce075c14aa9","last_reissued_at":"2026-07-05T09:50:47.529074Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:50:47.529074Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Quantifying Semantic Emergence in Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Hang Chen, Jiaying Zhu, Wenya Wang, Xinyu Yang","submitted_at":"2024-05-21T09:12:20Z","abstract_excerpt":"Large language models (LLMs) are widely recognized for their exceptional capacity to capture semantics meaning. Yet, there remains no established metric to quantify this capability. In this work, we introduce a quantitative metric, Information Emergence (IE), designed to measure LLMs' ability to extract semantics from input tokens. We formalize ``semantics'' as the meaningful information abstracted from a sequence of tokens and quantify this by comparing the entropy reduction observed for a sequence of tokens (macro-level) and individual tokens (micro-level). To achieve this, we design a light"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.12617","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/2405.12617/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":"2405.12617","created_at":"2026-07-05T09:50:47.529133+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.12617v2","created_at":"2026-07-05T09:50:47.529133+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.12617","created_at":"2026-07-05T09:50:47.529133+00:00"},{"alias_kind":"pith_short_12","alias_value":"NTGWAWUKTPFQ","created_at":"2026-07-05T09:50:47.529133+00:00"},{"alias_kind":"pith_short_16","alias_value":"NTGWAWUKTPFQDJAM","created_at":"2026-07-05T09:50:47.529133+00:00"},{"alias_kind":"pith_short_8","alias_value":"NTGWAWUK","created_at":"2026-07-05T09:50:47.529133+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2506.00688","citing_title":"Existing Large Language Model Unlearning Evaluations Are Inconclusive","ref_index":55,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY","json":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY.json","graph_json":"https://pith.science/api/pith-number/NTGWAWUKTPFQDJAM4WWGLQFIUY/graph.json","events_json":"https://pith.science/api/pith-number/NTGWAWUKTPFQDJAM4WWGLQFIUY/events.json","paper":"https://pith.science/paper/NTGWAWUK"},"agent_actions":{"view_html":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY","download_json":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY.json","view_paper":"https://pith.science/paper/NTGWAWUK","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.12617&json=true","fetch_graph":"https://pith.science/api/pith-number/NTGWAWUKTPFQDJAM4WWGLQFIUY/graph.json","fetch_events":"https://pith.science/api/pith-number/NTGWAWUKTPFQDJAM4WWGLQFIUY/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY/action/timestamp_anchor","attest_storage":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY/action/storage_attestation","attest_author":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY/action/author_attestation","sign_citation":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY/action/citation_signature","submit_replication":"https://pith.science/pith/NTGWAWUKTPFQDJAM4WWGLQFIUY/action/replication_record"}},"created_at":"2026-07-05T09:50:47.529133+00:00","updated_at":"2026-07-05T09:50:47.529133+00:00"}