{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:I23EX2FIKGT3S7L4DKA2R6P7DC","short_pith_number":"pith:I23EX2FI","schema_version":"1.0","canonical_sha256":"46b64be8a851a7b97d7c1a81a8f9ff189c39ee61c554443777ca8e2f4dbe2ca4","source":{"kind":"arxiv","id":"2408.08944","version":1},"attestation_state":"computed","paper":{"title":"Information-Theoretic Progress Measures reveal Grokking is an Emergent Phase Transition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IT","math.IT"],"primary_cat":"cs.LG","authors_text":"Daniele Marinazzo, Kenzo Clauw, Sebastiano Stramaglia","submitted_at":"2024-08-16T16:33:46Z","abstract_excerpt":"This paper studies emergent phenomena in neural networks by focusing on grokking where models suddenly generalize after delayed memorization. To understand this phase transition, we utilize higher-order mutual information to analyze the collective behavior (synergy) and shared properties (redundancy) between neurons during training. We identify distinct phases before grokking allowing us to anticipate when it occurs. We attribute grokking to an emergent phase transition caused by the synergistic interactions between neurons as a whole. We show that weight decay and weight initialization can en"},"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":"2408.08944","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2024-08-16T16:33:46Z","cross_cats_sorted":["cs.IT","math.IT"],"title_canon_sha256":"12ee0f4a1c20770e221e0b86c8fb9552a21fa037e42f553503352ee78051f49a","abstract_canon_sha256":"7246d29a2b9070242d3d048763fa39ae862fe85cab1d7d5642514e8afd63238f"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:56:25.932706Z","signature_b64":"ogUf9NkNEnkTCFkwG3zqtJ93vzDKvg8ZM98nflrCRTnPWu2jpeS8d8nc8ygg0dV75XoIsZxj/Uki+cO3/D6YBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"46b64be8a851a7b97d7c1a81a8f9ff189c39ee61c554443777ca8e2f4dbe2ca4","last_reissued_at":"2026-07-05T08:56:25.932302Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:56:25.932302Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Information-Theoretic Progress Measures reveal Grokking is an Emergent Phase Transition","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.IT","math.IT"],"primary_cat":"cs.LG","authors_text":"Daniele Marinazzo, Kenzo Clauw, Sebastiano Stramaglia","submitted_at":"2024-08-16T16:33:46Z","abstract_excerpt":"This paper studies emergent phenomena in neural networks by focusing on grokking where models suddenly generalize after delayed memorization. To understand this phase transition, we utilize higher-order mutual information to analyze the collective behavior (synergy) and shared properties (redundancy) between neurons during training. We identify distinct phases before grokking allowing us to anticipate when it occurs. We attribute grokking to an emergent phase transition caused by the synergistic interactions between neurons as a whole. We show that weight decay and weight initialization can en"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2408.08944","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/2408.08944/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":"2408.08944","created_at":"2026-07-05T08:56:25.932366+00:00"},{"alias_kind":"arxiv_version","alias_value":"2408.08944v1","created_at":"2026-07-05T08:56:25.932366+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2408.08944","created_at":"2026-07-05T08:56:25.932366+00:00"},{"alias_kind":"pith_short_12","alias_value":"I23EX2FIKGT3","created_at":"2026-07-05T08:56:25.932366+00:00"},{"alias_kind":"pith_short_16","alias_value":"I23EX2FIKGT3S7L4","created_at":"2026-07-05T08:56:25.932366+00:00"},{"alias_kind":"pith_short_8","alias_value":"I23EX2FI","created_at":"2026-07-05T08:56:25.932366+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":7,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.05863","citing_title":"Deciphering Two Training Clocks in Grokking via Deep Linear Network Theory with Conditional ReLU Reduction","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2606.30384","citing_title":"Scalar Representations of Neural Network Training Dynamics","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2606.07563","citing_title":"Emergence via Phase Transitions: Mechanism Landscapes and Universal Convergence Across Complex Systems","ref_index":33,"is_internal_anchor":false},{"citing_arxiv_id":"2605.16325","citing_title":"Phase Transitions in Driven Informational Systems: A Two-Field Perspective on Learning Theory and Non-Equilibrium Chemistry","ref_index":8,"is_internal_anchor":false},{"citing_arxiv_id":"2605.09724","citing_title":"Model Capacity Determines Grokking through Competing Memorisation and Generalisation Speeds","ref_index":32,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06352","citing_title":"Topological Signatures of Grokking","ref_index":5,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12434","citing_title":"A Bayesian Perspective on the Role of Epistemic Uncertainty for Delayed Generalization in In-Context Learning","ref_index":1,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC","json":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC.json","graph_json":"https://pith.science/api/pith-number/I23EX2FIKGT3S7L4DKA2R6P7DC/graph.json","events_json":"https://pith.science/api/pith-number/I23EX2FIKGT3S7L4DKA2R6P7DC/events.json","paper":"https://pith.science/paper/I23EX2FI"},"agent_actions":{"view_html":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC","download_json":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC.json","view_paper":"https://pith.science/paper/I23EX2FI","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2408.08944&json=true","fetch_graph":"https://pith.science/api/pith-number/I23EX2FIKGT3S7L4DKA2R6P7DC/graph.json","fetch_events":"https://pith.science/api/pith-number/I23EX2FIKGT3S7L4DKA2R6P7DC/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC/action/timestamp_anchor","attest_storage":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC/action/storage_attestation","attest_author":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC/action/author_attestation","sign_citation":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC/action/citation_signature","submit_replication":"https://pith.science/pith/I23EX2FIKGT3S7L4DKA2R6P7DC/action/replication_record"}},"created_at":"2026-07-05T08:56:25.932366+00:00","updated_at":"2026-07-05T08:56:25.932366+00:00"}