{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:P7GAEHENFGHATVOXCTR7O5EI4Z","short_pith_number":"pith:P7GAEHEN","schema_version":"1.0","canonical_sha256":"7fcc021c8d298e09d5d714e3f77488e643f589c82d54b4c0ad64cea404efcafe","source":{"kind":"arxiv","id":"2602.02304","version":2},"attestation_state":"computed","paper":{"title":"Comparing Explanations is Not Enough, Explain the Change: New Standards are Needed to Explain Behavioral Shifts in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Fosca Giannotti, Francesco Giannini, Luca Pappalardo, Martino Ciaperoni, Marzio Di Vece, Roberto Pellungrini","submitted_at":"2026-02-02T16:36:21Z","abstract_excerpt":"Large-scale foundation models exhibit \\emph{behavioral shifts} when subjected to interventions such as scaling, fine-tuning, reinforcement learning with human feedback, or in-context learning. Current explainability methods are structurally ill-suited to explain these shifts, because they either treat models as static objects, as traditional eXplainable AI (XAI) approaches do, or merely compare independent explanations across different checkpoints of a model. As a result, these approaches fail to explain the functional transition between two model instances in which a certain behavior has shif"},"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":"2602.02304","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2026-02-02T16:36:21Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"8b563dc2282582e984f299ff679f424fe7fd280c04f7a18caa1c51923d8fdd28","abstract_canon_sha256":"a0172cad6df8120c24ad825ea1fc81b3718b6cb733cb7f2aa2c9925a06cbec17"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:05:15.538148Z","signature_b64":"slMHnKkfFdDyLqkntlBRkX947ZQ+WmMm1XogTBo+UWq86pLg8tLBqEYFa8pKs1DTrlj+gR66yY023mDSs83OAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7fcc021c8d298e09d5d714e3f77488e643f589c82d54b4c0ad64cea404efcafe","last_reissued_at":"2026-05-21T01:05:15.537374Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:05:15.537374Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Comparing Explanations is Not Enough, Explain the Change: New Standards are Needed to Explain Behavioral Shifts in Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Fosca Giannotti, Francesco Giannini, Luca Pappalardo, Martino Ciaperoni, Marzio Di Vece, Roberto Pellungrini","submitted_at":"2026-02-02T16:36:21Z","abstract_excerpt":"Large-scale foundation models exhibit \\emph{behavioral shifts} when subjected to interventions such as scaling, fine-tuning, reinforcement learning with human feedback, or in-context learning. Current explainability methods are structurally ill-suited to explain these shifts, because they either treat models as static objects, as traditional eXplainable AI (XAI) approaches do, or merely compare independent explanations across different checkpoints of a model. As a result, these approaches fail to explain the functional transition between two model instances in which a certain behavior has shif"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2602.02304","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/2602.02304/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":"2602.02304","created_at":"2026-05-21T01:05:15.537509+00:00"},{"alias_kind":"arxiv_version","alias_value":"2602.02304v2","created_at":"2026-05-21T01:05:15.537509+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2602.02304","created_at":"2026-05-21T01:05:15.537509+00:00"},{"alias_kind":"pith_short_12","alias_value":"P7GAEHENFGHA","created_at":"2026-05-21T01:05:15.537509+00:00"},{"alias_kind":"pith_short_16","alias_value":"P7GAEHENFGHATVOX","created_at":"2026-05-21T01:05:15.537509+00:00"},{"alias_kind":"pith_short_8","alias_value":"P7GAEHEN","created_at":"2026-05-21T01:05:15.537509+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/P7GAEHENFGHATVOXCTR7O5EI4Z","json":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z.json","graph_json":"https://pith.science/api/pith-number/P7GAEHENFGHATVOXCTR7O5EI4Z/graph.json","events_json":"https://pith.science/api/pith-number/P7GAEHENFGHATVOXCTR7O5EI4Z/events.json","paper":"https://pith.science/paper/P7GAEHEN"},"agent_actions":{"view_html":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z","download_json":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z.json","view_paper":"https://pith.science/paper/P7GAEHEN","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2602.02304&json=true","fetch_graph":"https://pith.science/api/pith-number/P7GAEHENFGHATVOXCTR7O5EI4Z/graph.json","fetch_events":"https://pith.science/api/pith-number/P7GAEHENFGHATVOXCTR7O5EI4Z/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z/action/timestamp_anchor","attest_storage":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z/action/storage_attestation","attest_author":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z/action/author_attestation","sign_citation":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z/action/citation_signature","submit_replication":"https://pith.science/pith/P7GAEHENFGHATVOXCTR7O5EI4Z/action/replication_record"}},"created_at":"2026-05-21T01:05:15.537509+00:00","updated_at":"2026-05-21T01:05:15.537509+00:00"}