{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:O2FT256RNL4QQ4DC5DT2JEVNCZ","short_pith_number":"pith:O2FT256R","schema_version":"1.0","canonical_sha256":"768b3d77d16af9087062e8e7a492ad166c67bd98f23793f3f239fe8564d6e5f5","source":{"kind":"arxiv","id":"2506.13060","version":1},"attestation_state":"computed","paper":{"title":"Rethinking Explainability in the Era of Multimodal AI","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Chirag Agarwal","submitted_at":"2025-06-16T03:08:29Z","abstract_excerpt":"While multimodal AI systems (models jointly trained on heterogeneous data types such as text, time series, graphs, and images) have become ubiquitous and achieved remarkable performance across high-stakes applications, transparent and accurate explanation algorithms are crucial for their safe deployment and ensure user trust. However, most existing explainability techniques remain unimodal, generating modality-specific feature attributions, concepts, or circuit traces in isolation and thus failing to capture cross-modal interactions. This paper argues that such unimodal explanations systematic"},"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":"2506.13060","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-06-16T03:08:29Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"cdcc1d924a56a7d50dd67a1d9c579247bf9fc1a05c23c98e9907774a09dcea5a","abstract_canon_sha256":"707cf158937201b15134416c2fd53c2a4ec495b37235045b92af1b5f72332f38"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:22:04.509722Z","signature_b64":"IVd8w/wISHiVWmD9Zk5OY1DOvuGIn2QMNFv0Ly0ybuMKXuPgF0faOQgG+VP/SdmuLGu2pW2FoDApb9a8aHydBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"768b3d77d16af9087062e8e7a492ad166c67bd98f23793f3f239fe8564d6e5f5","last_reissued_at":"2026-07-05T11:22:04.509326Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:22:04.509326Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Rethinking Explainability in the Era of Multimodal AI","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.AI","authors_text":"Chirag Agarwal","submitted_at":"2025-06-16T03:08:29Z","abstract_excerpt":"While multimodal AI systems (models jointly trained on heterogeneous data types such as text, time series, graphs, and images) have become ubiquitous and achieved remarkable performance across high-stakes applications, transparent and accurate explanation algorithms are crucial for their safe deployment and ensure user trust. However, most existing explainability techniques remain unimodal, generating modality-specific feature attributions, concepts, or circuit traces in isolation and thus failing to capture cross-modal interactions. This paper argues that such unimodal explanations systematic"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2506.13060","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/2506.13060/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":"2506.13060","created_at":"2026-07-05T11:22:04.509386+00:00"},{"alias_kind":"arxiv_version","alias_value":"2506.13060v1","created_at":"2026-07-05T11:22:04.509386+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2506.13060","created_at":"2026-07-05T11:22:04.509386+00:00"},{"alias_kind":"pith_short_12","alias_value":"O2FT256RNL4Q","created_at":"2026-07-05T11:22:04.509386+00:00"},{"alias_kind":"pith_short_16","alias_value":"O2FT256RNL4QQ4DC","created_at":"2026-07-05T11:22:04.509386+00:00"},{"alias_kind":"pith_short_8","alias_value":"O2FT256R","created_at":"2026-07-05T11:22:04.509386+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.03712","citing_title":"When Graph Tokens Sink: A Mechanistic Analysis of Graph Language Models","ref_index":25,"is_internal_anchor":false},{"citing_arxiv_id":"2605.22168","citing_title":"Measuring Cross-Modal Synergy: A Benchmark for VLM Explainability","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2508.04427","citing_title":"Decoding the Multimodal Maze: A Systematic Review on the Adoption of Explainability in Multimodal Attention-based Models","ref_index":132,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ","json":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ.json","graph_json":"https://pith.science/api/pith-number/O2FT256RNL4QQ4DC5DT2JEVNCZ/graph.json","events_json":"https://pith.science/api/pith-number/O2FT256RNL4QQ4DC5DT2JEVNCZ/events.json","paper":"https://pith.science/paper/O2FT256R"},"agent_actions":{"view_html":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ","download_json":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ.json","view_paper":"https://pith.science/paper/O2FT256R","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2506.13060&json=true","fetch_graph":"https://pith.science/api/pith-number/O2FT256RNL4QQ4DC5DT2JEVNCZ/graph.json","fetch_events":"https://pith.science/api/pith-number/O2FT256RNL4QQ4DC5DT2JEVNCZ/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ/action/timestamp_anchor","attest_storage":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ/action/storage_attestation","attest_author":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ/action/author_attestation","sign_citation":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ/action/citation_signature","submit_replication":"https://pith.science/pith/O2FT256RNL4QQ4DC5DT2JEVNCZ/action/replication_record"}},"created_at":"2026-07-05T11:22:04.509386+00:00","updated_at":"2026-07-05T11:22:04.509386+00:00"}