{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2022:Z5R5YYA3AIYJZKI7CNVJQ5HMGL","short_pith_number":"pith:Z5R5YYA3","schema_version":"1.0","canonical_sha256":"cf63dc601b02309ca91f136a9874ec32dc28ab4ccb298bb903392fdfd8a6dc76","source":{"kind":"arxiv","id":"2205.15465","version":1},"attestation_state":"computed","paper":{"title":"Analyzing Modality Robustness in Multimodal Sentiment Analysis","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bo Cheng, Devamanyu Hazarika, Roger Zimmermann, Shuai Zhao, Soujanya Poria, Yingting Li","submitted_at":"2022-05-30T23:30:16Z","abstract_excerpt":"Building robust multimodal models are crucial for achieving reliable deployment in the wild. Despite its importance, less attention has been paid to identifying and improving the robustness of Multimodal Sentiment Analysis (MSA) models. In this work, we hope to address that by (i) Proposing simple diagnostic checks for modality robustness in a trained multimodal model. Using these checks, we find MSA models to be highly sensitive to a single modality, which creates issues in their robustness; (ii) We analyze well-known robust training strategies to alleviate the issues. Critically, we observe "},"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":"2205.15465","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2022-05-30T23:30:16Z","cross_cats_sorted":[],"title_canon_sha256":"0f3b6aee7731c09e3d6d944f5aa60dc542fabb5e1b051a21073bba86b1aeeb84","abstract_canon_sha256":"7eb4bd1be30e88a026bfc8fa85faf7fa819d740b4b33be14d45e852cd903a106"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:27:33.120615Z","signature_b64":"yxEkMmKi4150qQXKRtdOmWonHGifIwAbGY3DWRo11p7Qq1yLraUDca73pn31EIDo6CcOGmVUwMKX24X4Cc/yDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cf63dc601b02309ca91f136a9874ec32dc28ab4ccb298bb903392fdfd8a6dc76","last_reissued_at":"2026-07-05T04:27:33.120077Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:27:33.120077Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Analyzing Modality Robustness in Multimodal Sentiment Analysis","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bo Cheng, Devamanyu Hazarika, Roger Zimmermann, Shuai Zhao, Soujanya Poria, Yingting Li","submitted_at":"2022-05-30T23:30:16Z","abstract_excerpt":"Building robust multimodal models are crucial for achieving reliable deployment in the wild. Despite its importance, less attention has been paid to identifying and improving the robustness of Multimodal Sentiment Analysis (MSA) models. In this work, we hope to address that by (i) Proposing simple diagnostic checks for modality robustness in a trained multimodal model. Using these checks, we find MSA models to be highly sensitive to a single modality, which creates issues in their robustness; (ii) We analyze well-known robust training strategies to alleviate the issues. Critically, we observe "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2205.15465","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/2205.15465/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":"2205.15465","created_at":"2026-07-05T04:27:33.120144+00:00"},{"alias_kind":"arxiv_version","alias_value":"2205.15465v1","created_at":"2026-07-05T04:27:33.120144+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2205.15465","created_at":"2026-07-05T04:27:33.120144+00:00"},{"alias_kind":"pith_short_12","alias_value":"Z5R5YYA3AIYJ","created_at":"2026-07-05T04:27:33.120144+00:00"},{"alias_kind":"pith_short_16","alias_value":"Z5R5YYA3AIYJZKI7","created_at":"2026-07-05T04:27:33.120144+00:00"},{"alias_kind":"pith_short_8","alias_value":"Z5R5YYA3","created_at":"2026-07-05T04:27:33.120144+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2602.16161","citing_title":"Emotion Collider: Dual Hyperbolic Mirror Manifolds for Sentiment Recovery via Anti Emotion Reflection","ref_index":25,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL","json":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL.json","graph_json":"https://pith.science/api/pith-number/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/graph.json","events_json":"https://pith.science/api/pith-number/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/events.json","paper":"https://pith.science/paper/Z5R5YYA3"},"agent_actions":{"view_html":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL","download_json":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL.json","view_paper":"https://pith.science/paper/Z5R5YYA3","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2205.15465&json=true","fetch_graph":"https://pith.science/api/pith-number/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/graph.json","fetch_events":"https://pith.science/api/pith-number/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/action/storage_attestation","attest_author":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/action/author_attestation","sign_citation":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/action/citation_signature","submit_replication":"https://pith.science/pith/Z5R5YYA3AIYJZKI7CNVJQ5HMGL/action/replication_record"}},"created_at":"2026-07-05T04:27:33.120144+00:00","updated_at":"2026-07-05T04:27:33.120144+00:00"}