{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:45LPRMJ7TTQRWKNQYIZRHWEBZI","short_pith_number":"pith:45LPRMJ7","canonical_record":{"source":{"id":"2405.18654","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T23:36:00Z","cross_cats_sorted":[],"title_canon_sha256":"2d63bc166586c766c0f90dbc4af350e8d4fd71a739a1d8099ce41f04d741ff78","abstract_canon_sha256":"61fe9ca4c537eaf8b097258a0f236cdb9df32c50b0a3c057ee7bcf8b55e530b3"},"schema_version":"1.0"},"canonical_sha256":"e756f8b13f9ce11b29b0c23313d881ca0fd51f156a763736d4348afee5de8479","source":{"kind":"arxiv","id":"2405.18654","version":3},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.18654","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"arxiv_version","alias_value":"2405.18654v3","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18654","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_12","alias_value":"45LPRMJ7TTQR","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_16","alias_value":"45LPRMJ7TTQRWKNQ","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_8","alias_value":"45LPRMJ7","created_at":"2026-07-05T10:21:19Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:45LPRMJ7TTQRWKNQYIZRHWEBZI","target":"record","payload":{"canonical_record":{"source":{"id":"2405.18654","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T23:36:00Z","cross_cats_sorted":[],"title_canon_sha256":"2d63bc166586c766c0f90dbc4af350e8d4fd71a739a1d8099ce41f04d741ff78","abstract_canon_sha256":"61fe9ca4c537eaf8b097258a0f236cdb9df32c50b0a3c057ee7bcf8b55e530b3"},"schema_version":"1.0"},"canonical_sha256":"e756f8b13f9ce11b29b0c23313d881ca0fd51f156a763736d4348afee5de8479","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:21:19.609091Z","signature_b64":"vCVzvV6gfQSxr3p3zmchS7vuqXAMMPCmRnk4vmfZdgWtVwbxc8mL54DM4OqDdaZNSl02pUQxeg8bebteYurODg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"e756f8b13f9ce11b29b0c23313d881ca0fd51f156a763736d4348afee5de8479","last_reissued_at":"2026-07-05T10:21:19.608600Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:21:19.608600Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.18654","source_version":3,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:21:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"W698b1ZlNzlcH4UD4NMVd7xBDBO36bcJYQIX7Re1shL9VjBfQe/DWODF9DxpmcLFY3MQihtYFM1Oh57TUnwuCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:25:33.213881Z"},"content_sha256":"91509694d6f4a1e9360ee721d5e65a476bed35fbf0acd53934b9b1078d0b9945","schema_version":"1.0","event_id":"sha256:91509694d6f4a1e9360ee721d5e65a476bed35fbf0acd53934b9b1078d0b9945"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:45LPRMJ7TTQRWKNQYIZRHWEBZI","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Mitigating Object Hallucination in MLLMs via Data-augmented Phrase-level Alignment","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Ahmad Beirami, Ali Etemad, Pritam Sarkar, Sayna Ebrahimi, Sercan \\\"O. Ar{\\i}k, Tomas Pfister","submitted_at":"2024-05-28T23:36:00Z","abstract_excerpt":"Despite their significant advancements, Multimodal Large Language Models (MLLMs) often generate factually inaccurate information, referred to as hallucination. In this work, we address object hallucinations in MLLMs, where information is generated about an object not present in the input image. We introduce Data-augmented Phrase-level Alignment (DPA), a novel loss which can be applied to instruction-tuned off-the-shelf MLLMs to mitigate hallucinations, while preserving their general vision-language capabilities. To fine-tune MLLMs with DPA, we first generate a set of `hallucinated' and `correc"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18654","kind":"arxiv","version":3},"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/2405.18654/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":"2026-07-05T10:21:19Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uhG84lxV5SXOEKAhg+l3ERiRsLNUizX5zN6rFqV0h34Rcr2TQ12FHqzc9KqE6XX8fV3I7VSyIj1XQwmr+r6EAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-08T06:25:33.214368Z"},"content_sha256":"012995ce8857a5a47f74393eb561c869b647140c034fd066e7c211d2ad8c8dd2","schema_version":"1.0","event_id":"sha256:012995ce8857a5a47f74393eb561c869b647140c034fd066e7c211d2ad8c8dd2"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/bundle.json","state_url":"https://pith.science/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/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":"2026-08-08T06:25:33Z","links":{"resolver":"https://pith.science/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI","bundle":"https://pith.science/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/bundle.json","state":"https://pith.science/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/state.json","well_known_bundle":"https://pith.science/.well-known/pith/45LPRMJ7TTQRWKNQYIZRHWEBZI/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:45LPRMJ7TTQRWKNQYIZRHWEBZI","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":"61fe9ca4c537eaf8b097258a0f236cdb9df32c50b0a3c057ee7bcf8b55e530b3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T23:36:00Z","title_canon_sha256":"2d63bc166586c766c0f90dbc4af350e8d4fd71a739a1d8099ce41f04d741ff78"},"schema_version":"1.0","source":{"id":"2405.18654","kind":"arxiv","version":3}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.18654","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"arxiv_version","alias_value":"2405.18654v3","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.18654","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_12","alias_value":"45LPRMJ7TTQR","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_16","alias_value":"45LPRMJ7TTQRWKNQ","created_at":"2026-07-05T10:21:19Z"},{"alias_kind":"pith_short_8","alias_value":"45LPRMJ7","created_at":"2026-07-05T10:21:19Z"}],"graph_snapshots":[{"event_id":"sha256:012995ce8857a5a47f74393eb561c869b647140c034fd066e7c211d2ad8c8dd2","target":"graph","created_at":"2026-07-05T10:21:19Z","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/2405.18654/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Despite their significant advancements, Multimodal Large Language Models (MLLMs) often generate factually inaccurate information, referred to as hallucination. In this work, we address object hallucinations in MLLMs, where information is generated about an object not present in the input image. We introduce Data-augmented Phrase-level Alignment (DPA), a novel loss which can be applied to instruction-tuned off-the-shelf MLLMs to mitigate hallucinations, while preserving their general vision-language capabilities. To fine-tune MLLMs with DPA, we first generate a set of `hallucinated' and `correc","authors_text":"Ahmad Beirami, Ali Etemad, Pritam Sarkar, Sayna Ebrahimi, Sercan \\\"O. Ar{\\i}k, Tomas Pfister","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T23:36:00Z","title":"Mitigating Object Hallucination in MLLMs via Data-augmented Phrase-level Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.18654","kind":"arxiv","version":3},"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":"sha256:91509694d6f4a1e9360ee721d5e65a476bed35fbf0acd53934b9b1078d0b9945","target":"record","created_at":"2026-07-05T10:21:19Z","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":"61fe9ca4c537eaf8b097258a0f236cdb9df32c50b0a3c057ee7bcf8b55e530b3","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-28T23:36:00Z","title_canon_sha256":"2d63bc166586c766c0f90dbc4af350e8d4fd71a739a1d8099ce41f04d741ff78"},"schema_version":"1.0","source":{"id":"2405.18654","kind":"arxiv","version":3}},"canonical_sha256":"e756f8b13f9ce11b29b0c23313d881ca0fd51f156a763736d4348afee5de8479","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"e756f8b13f9ce11b29b0c23313d881ca0fd51f156a763736d4348afee5de8479","first_computed_at":"2026-07-05T10:21:19.608600Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:21:19.608600Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"vCVzvV6gfQSxr3p3zmchS7vuqXAMMPCmRnk4vmfZdgWtVwbxc8mL54DM4OqDdaZNSl02pUQxeg8bebteYurODg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:21:19.609091Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.18654","source_kind":"arxiv","source_version":3}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:91509694d6f4a1e9360ee721d5e65a476bed35fbf0acd53934b9b1078d0b9945","sha256:012995ce8857a5a47f74393eb561c869b647140c034fd066e7c211d2ad8c8dd2"],"state_sha256":"995e969c419dcc8fceecfa5d0f7c2c82f20752175127e712552995bab8755969"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ghDJwCf1BaQcUcvtLc8++KWAxzQUiogY4SHP0RB5iZpn1x5IbrEJGUM8rDvWC7h7JwgpqE0mKjswPc6Q8H1pAA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-08T06:25:33.217855Z","bundle_sha256":"ac2d303acce12ff1c9874c9f6971dfaa3199b05041d806734cc36a03ed3de89d"}}