{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:FSTZXMWN7ZG7GJKB2A7MT4RVKQ","short_pith_number":"pith:FSTZXMWN","schema_version":"1.0","canonical_sha256":"2ca79bb2cdfe4df32541d03ec9f235543078ffdcb3ad53372b4c871331889479","source":{"kind":"arxiv","id":"2411.16035","version":1},"attestation_state":"computed","paper":{"title":"Predicting Emergent Capabilities by Finetuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Charlie Snell, Dan Klein, Eric Wallace, Sergey Levine","submitted_at":"2024-11-25T01:48:09Z","abstract_excerpt":"A fundamental open challenge in modern LLM scaling is the lack of understanding around emergent capabilities. In particular, language model pretraining loss is known to be highly predictable as a function of compute. However, downstream capabilities are far less predictable -- sometimes even exhibiting emergent jumps -- which makes it challenging to anticipate the capabilities of future models. In this work, we first pose the task of emergence prediction: given access to current LLMs that have random few-shot accuracy on a task, can we predict whether future models (GPT-N+1) will have non-triv"},"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":"2411.16035","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-11-25T01:48:09Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"6a4534219aad93b2b53a3690bb6137a973c2db2d52d232fd76eb555f83aa0284","abstract_canon_sha256":"d8762715a64bee1b63b9d6105e36736f7b150bef8e21e54f25c5897b16170663"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:40:02.901308Z","signature_b64":"yHCzUgrUJAu8Xs7m+ZBJuQkwpCXSHRufgeJ7BOex/JQ8+wPSqboXB6pW+lm+Rb5ZR35rJ7l+c/FgWK2uO4B4CA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"2ca79bb2cdfe4df32541d03ec9f235543078ffdcb3ad53372b4c871331889479","last_reissued_at":"2026-07-05T09:40:02.900848Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:40:02.900848Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Predicting Emergent Capabilities by Finetuning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Charlie Snell, Dan Klein, Eric Wallace, Sergey Levine","submitted_at":"2024-11-25T01:48:09Z","abstract_excerpt":"A fundamental open challenge in modern LLM scaling is the lack of understanding around emergent capabilities. In particular, language model pretraining loss is known to be highly predictable as a function of compute. However, downstream capabilities are far less predictable -- sometimes even exhibiting emergent jumps -- which makes it challenging to anticipate the capabilities of future models. In this work, we first pose the task of emergence prediction: given access to current LLMs that have random few-shot accuracy on a task, can we predict whether future models (GPT-N+1) will have non-triv"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2411.16035","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/2411.16035/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":"2411.16035","created_at":"2026-07-05T09:40:02.900900+00:00"},{"alias_kind":"arxiv_version","alias_value":"2411.16035v1","created_at":"2026-07-05T09:40:02.900900+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2411.16035","created_at":"2026-07-05T09:40:02.900900+00:00"},{"alias_kind":"pith_short_12","alias_value":"FSTZXMWN7ZG7","created_at":"2026-07-05T09:40:02.900900+00:00"},{"alias_kind":"pith_short_16","alias_value":"FSTZXMWN7ZG7GJKB","created_at":"2026-07-05T09:40:02.900900+00:00"},{"alias_kind":"pith_short_8","alias_value":"FSTZXMWN","created_at":"2026-07-05T09:40:02.900900+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/FSTZXMWN7ZG7GJKB2A7MT4RVKQ","json":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ.json","graph_json":"https://pith.science/api/pith-number/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/graph.json","events_json":"https://pith.science/api/pith-number/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/events.json","paper":"https://pith.science/paper/FSTZXMWN"},"agent_actions":{"view_html":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ","download_json":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ.json","view_paper":"https://pith.science/paper/FSTZXMWN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2411.16035&json=true","fetch_graph":"https://pith.science/api/pith-number/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/graph.json","fetch_events":"https://pith.science/api/pith-number/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/action/storage_attestation","attest_author":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/action/author_attestation","sign_citation":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/action/citation_signature","submit_replication":"https://pith.science/pith/FSTZXMWN7ZG7GJKB2A7MT4RVKQ/action/replication_record"}},"created_at":"2026-07-05T09:40:02.900900+00:00","updated_at":"2026-07-05T09:40:02.900900+00:00"}