{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:HSYI2ULGVPPYOLSKGFFDWYIFPB","short_pith_number":"pith:HSYI2ULG","canonical_record":{"source":{"id":"2606.00230","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T18:04:52Z","cross_cats_sorted":[],"title_canon_sha256":"2660ac98a070ffa772de9be464f20f5e4e90d813394c7c320a9dc0476f71c3ab","abstract_canon_sha256":"3ad19bd65075f393620aa73412184226abc1f2efa4dfa6d5181b84996189e460"},"schema_version":"1.0"},"canonical_sha256":"3cb08d5166abdf872e4a314a3b61057848e7dd3bf5b11238182a2b6fc65df2bc","source":{"kind":"arxiv","id":"2606.00230","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.00230","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"arxiv_version","alias_value":"2606.00230v1","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00230","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_12","alias_value":"HSYI2ULGVPPY","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_16","alias_value":"HSYI2ULGVPPYOLSK","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_8","alias_value":"HSYI2ULG","created_at":"2026-06-02T01:03:22Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:HSYI2ULGVPPYOLSKGFFDWYIFPB","target":"record","payload":{"canonical_record":{"source":{"id":"2606.00230","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T18:04:52Z","cross_cats_sorted":[],"title_canon_sha256":"2660ac98a070ffa772de9be464f20f5e4e90d813394c7c320a9dc0476f71c3ab","abstract_canon_sha256":"3ad19bd65075f393620aa73412184226abc1f2efa4dfa6d5181b84996189e460"},"schema_version":"1.0"},"canonical_sha256":"3cb08d5166abdf872e4a314a3b61057848e7dd3bf5b11238182a2b6fc65df2bc","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-02T01:03:22.303434Z","signature_b64":"8kwdNGLf1SGz2BNo/2HuLeKY5iTlitVhFzRPz0dEdH+FFN1/VUb8+Yi7201odPXi+ly3hblLMnd3eUcoAJSeDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"3cb08d5166abdf872e4a314a3b61057848e7dd3bf5b11238182a2b6fc65df2bc","last_reissued_at":"2026-06-02T01:03:22.303064Z","signature_status":"signed_v1","first_computed_at":"2026-06-02T01:03:22.303064Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.00230","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T01:03:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"hXINSP2c4m82kNnv8oNugMqFwYqf16b0TOuQ5QU7n5jls9RnJ+covB5AOrZdfuUMAWWR83XM1XIHeMG7iTHAAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T11:55:58.304723Z"},"content_sha256":"2eb57a3930c351480e976a57059163828086442362e9fe282bbda8c0b4a436b9","schema_version":"1.0","event_id":"sha256:2eb57a3930c351480e976a57059163828086442362e9fe282bbda8c0b4a436b9"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:HSYI2ULGVPPYOLSKGFFDWYIFPB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"A Pre-Training Analogue of Grokking in Language Models: Tracing Delayed Grammatical Generalization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Anna Rumshisky, Namrata Shivagunde, Sherin Muckatira, Vijeta Deshpande","submitted_at":"2026-05-29T18:04:52Z","abstract_excerpt":"Grokking, the phenomenon in which neural networks generalize long after fitting their training data, has been studied in supervised settings on many epochs. LLM pre-training instead involves next-token prediction over an unlabeled corpus, with limited data repetition and no explicit train/validation split. To address this, we propose an exposure-based framework that enables the study of grokking-like dynamics during LLM pre-training. We ground our evaluation in BLiMP minimal pairs, which provide controlled grammatical contrasts. For every BLiMP minimal pair, we identify a critical phrase, the "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00230","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/2606.00230/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-06-02T01:03:22Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"18ONtHTl6Py48gKgzIMNzBF86MnFoF8e5CSPqGFQu7guehDNN9Xjw+Anh/EC/GiPA35MuzuFu/xoHCpxwkQxDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-29T11:55:58.305103Z"},"content_sha256":"e226b60bca3e8fd2982067487352f3cc53cb1312333ac70340197e2dff67d6e2","schema_version":"1.0","event_id":"sha256:e226b60bca3e8fd2982067487352f3cc53cb1312333ac70340197e2dff67d6e2"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/bundle.json","state_url":"https://pith.science/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-06-29T11:55:58Z","links":{"resolver":"https://pith.science/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB","bundle":"https://pith.science/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/bundle.json","state":"https://pith.science/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/HSYI2ULGVPPYOLSKGFFDWYIFPB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:HSYI2ULGVPPYOLSKGFFDWYIFPB","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"3ad19bd65075f393620aa73412184226abc1f2efa4dfa6d5181b84996189e460","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T18:04:52Z","title_canon_sha256":"2660ac98a070ffa772de9be464f20f5e4e90d813394c7c320a9dc0476f71c3ab"},"schema_version":"1.0","source":{"id":"2606.00230","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.00230","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"arxiv_version","alias_value":"2606.00230v1","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.00230","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_12","alias_value":"HSYI2ULGVPPY","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_16","alias_value":"HSYI2ULGVPPYOLSK","created_at":"2026-06-02T01:03:22Z"},{"alias_kind":"pith_short_8","alias_value":"HSYI2ULG","created_at":"2026-06-02T01:03:22Z"}],"graph_snapshots":[{"event_id":"sha256:e226b60bca3e8fd2982067487352f3cc53cb1312333ac70340197e2dff67d6e2","target":"graph","created_at":"2026-06-02T01:03:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2606.00230/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Grokking, the phenomenon in which neural networks generalize long after fitting their training data, has been studied in supervised settings on many epochs. LLM pre-training instead involves next-token prediction over an unlabeled corpus, with limited data repetition and no explicit train/validation split. To address this, we propose an exposure-based framework that enables the study of grokking-like dynamics during LLM pre-training. We ground our evaluation in BLiMP minimal pairs, which provide controlled grammatical contrasts. For every BLiMP minimal pair, we identify a critical phrase, the ","authors_text":"Anna Rumshisky, Namrata Shivagunde, Sherin Muckatira, Vijeta Deshpande","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T18:04:52Z","title":"A Pre-Training Analogue of Grokking in Language Models: Tracing Delayed Grammatical Generalization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.00230","kind":"arxiv","version":1},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:2eb57a3930c351480e976a57059163828086442362e9fe282bbda8c0b4a436b9","target":"record","created_at":"2026-06-02T01:03:22Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"3ad19bd65075f393620aa73412184226abc1f2efa4dfa6d5181b84996189e460","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2026-05-29T18:04:52Z","title_canon_sha256":"2660ac98a070ffa772de9be464f20f5e4e90d813394c7c320a9dc0476f71c3ab"},"schema_version":"1.0","source":{"id":"2606.00230","kind":"arxiv","version":1}},"canonical_sha256":"3cb08d5166abdf872e4a314a3b61057848e7dd3bf5b11238182a2b6fc65df2bc","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"3cb08d5166abdf872e4a314a3b61057848e7dd3bf5b11238182a2b6fc65df2bc","first_computed_at":"2026-06-02T01:03:22.303064Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-02T01:03:22.303064Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"8kwdNGLf1SGz2BNo/2HuLeKY5iTlitVhFzRPz0dEdH+FFN1/VUb8+Yi7201odPXi+ly3hblLMnd3eUcoAJSeDA==","signature_status":"signed_v1","signed_at":"2026-06-02T01:03:22.303434Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.00230","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:2eb57a3930c351480e976a57059163828086442362e9fe282bbda8c0b4a436b9","sha256:e226b60bca3e8fd2982067487352f3cc53cb1312333ac70340197e2dff67d6e2"],"state_sha256":"81648a8d7dafda5e1bdb37b3d6048dd8b1c5748f15f55a8d4fc9e0ca79787167"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"t1eD8bnoQ9POmg5zeCIFXDZdJoLpTKmJWgX1q+40k1q58LHzJAMevOKh/rv5u4+X7eE/yTQpgWB9tnYeE/H9AA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-29T11:55:58.307361Z","bundle_sha256":"02d996eb07803a5825f2afaa85d3304af99d3b3523d24dd12d33458c7cb6b2da"}}