{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:VJSXRF4FGJQQDK4UYSNUD3MXP4","short_pith_number":"pith:VJSXRF4F","schema_version":"1.0","canonical_sha256":"aa65789785326101ab94c49b41ed977f12b4a2738251dafd0b3e91e75a101dea","source":{"kind":"arxiv","id":"2605.28826","version":1},"attestation_state":"computed","paper":{"title":"From Context Shift to Stylistic Collapse: Why Training Objectives Matter More Than Scale","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Rohan Mahapatra","submitted_at":"2026-04-08T02:13:46Z","abstract_excerpt":"In modern LLMs, linguistic features function not as stylistic artifacts but as probes of probability mass, allocated under training alignment objectives. Language models trained with contemporary pipelines exhibit severe reshaping of linguistic features, leading to extreme language re-distribution. While previous stylometric analyses explored linguistic differences between AI-generated and human texts, we focus on the reshaping plaguing the LLM training pipeline itself. We analyze 17 models (410M-100B+ parameters) across 24 linguistically-motivated probes, documenting that instruction-tuned sy"},"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":"2605.28826","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2026-04-08T02:13:46Z","cross_cats_sorted":[],"title_canon_sha256":"872e4bbef8903507c865b7e145e5221315fc613b4d16c38a754c410db19203ff","abstract_canon_sha256":"c485eacf556a82200d440ab8f7040390705345de42536b11af34c0d617b3d2d8"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-29T00:04:13.077259Z","signature_b64":"sQnMo3yYOyLAxbHeGjxHq63hJfL0xBA0KJgn9P6COOZL2BP1f2PBhjI4gKNLoNqqxIE8kW3egPPlQWWsKYTWAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"aa65789785326101ab94c49b41ed977f12b4a2738251dafd0b3e91e75a101dea","last_reissued_at":"2026-05-29T00:04:13.076671Z","signature_status":"signed_v1","first_computed_at":"2026-05-29T00:04:13.076671Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"From Context Shift to Stylistic Collapse: Why Training Objectives Matter More Than Scale","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Rohan Mahapatra","submitted_at":"2026-04-08T02:13:46Z","abstract_excerpt":"In modern LLMs, linguistic features function not as stylistic artifacts but as probes of probability mass, allocated under training alignment objectives. Language models trained with contemporary pipelines exhibit severe reshaping of linguistic features, leading to extreme language re-distribution. While previous stylometric analyses explored linguistic differences between AI-generated and human texts, we focus on the reshaping plaguing the LLM training pipeline itself. We analyze 17 models (410M-100B+ parameters) across 24 linguistically-motivated probes, documenting that instruction-tuned sy"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.28826","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/2605.28826/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":"2605.28826","created_at":"2026-05-29T00:04:13.076741+00:00"},{"alias_kind":"arxiv_version","alias_value":"2605.28826v1","created_at":"2026-05-29T00:04:13.076741+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.28826","created_at":"2026-05-29T00:04:13.076741+00:00"},{"alias_kind":"pith_short_12","alias_value":"VJSXRF4FGJQQ","created_at":"2026-05-29T00:04:13.076741+00:00"},{"alias_kind":"pith_short_16","alias_value":"VJSXRF4FGJQQDK4U","created_at":"2026-05-29T00:04:13.076741+00:00"},{"alias_kind":"pith_short_8","alias_value":"VJSXRF4F","created_at":"2026-05-29T00:04:13.076741+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/VJSXRF4FGJQQDK4UYSNUD3MXP4","json":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4.json","graph_json":"https://pith.science/api/pith-number/VJSXRF4FGJQQDK4UYSNUD3MXP4/graph.json","events_json":"https://pith.science/api/pith-number/VJSXRF4FGJQQDK4UYSNUD3MXP4/events.json","paper":"https://pith.science/paper/VJSXRF4F"},"agent_actions":{"view_html":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4","download_json":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4.json","view_paper":"https://pith.science/paper/VJSXRF4F","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2605.28826&json=true","fetch_graph":"https://pith.science/api/pith-number/VJSXRF4FGJQQDK4UYSNUD3MXP4/graph.json","fetch_events":"https://pith.science/api/pith-number/VJSXRF4FGJQQDK4UYSNUD3MXP4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4/action/storage_attestation","attest_author":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4/action/author_attestation","sign_citation":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4/action/citation_signature","submit_replication":"https://pith.science/pith/VJSXRF4FGJQQDK4UYSNUD3MXP4/action/replication_record"}},"created_at":"2026-05-29T00:04:13.076741+00:00","updated_at":"2026-05-29T00:04:13.076741+00:00"}