{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:U6EDSYK3M5P3QWF4YUEGU3QMZ5","short_pith_number":"pith:U6EDSYK3","canonical_record":{"source":{"id":"2503.06486","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T07:07:03Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4407fee008cbdb0ea314bfdef3767463c8de1c5ec367263381356634f5ee0b91","abstract_canon_sha256":"04540f22d8d0e341dfee01da002d7d2c279e8804a1b1df07d00f1c5c4e2c1da2"},"schema_version":"1.0"},"canonical_sha256":"a78839615b675fb858bcc5086a6e0ccf7059f10e99d8f355825e32bc520ebc16","source":{"kind":"arxiv","id":"2503.06486","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.06486","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"arxiv_version","alias_value":"2503.06486v1","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06486","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_12","alias_value":"U6EDSYK3M5P3","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_16","alias_value":"U6EDSYK3M5P3QWF4","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_8","alias_value":"U6EDSYK3","created_at":"2026-07-05T10:27:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:U6EDSYK3M5P3QWF4YUEGU3QMZ5","target":"record","payload":{"canonical_record":{"source":{"id":"2503.06486","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T07:07:03Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"4407fee008cbdb0ea314bfdef3767463c8de1c5ec367263381356634f5ee0b91","abstract_canon_sha256":"04540f22d8d0e341dfee01da002d7d2c279e8804a1b1df07d00f1c5c4e2c1da2"},"schema_version":"1.0"},"canonical_sha256":"a78839615b675fb858bcc5086a6e0ccf7059f10e99d8f355825e32bc520ebc16","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:27:35.079824Z","signature_b64":"HELCyf8rDaWgllxmUxryrNNxMMVl8l/3h0iiD1g4KGjR8TFcr5A3kbWhvPBiHoWN0TtS3w58EyuXva+P3PScBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a78839615b675fb858bcc5086a6e0ccf7059f10e99d8f355825e32bc520ebc16","last_reissued_at":"2026-07-05T10:27:35.079188Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:27:35.079188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.06486","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T10:27:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"NWIW6n0FvtIyfXVumCOR0tDpD82WR7H8e0YtsISgGjt2rYOkLdjqrTD/5SuWGlPy+1VcxpxCscbMxlCFjyWZBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T05:22:57.068140Z"},"content_sha256":"b402e8d018701d34f228510bae3239faca723ffc1357fd6f34928d9a99f64f48","schema_version":"1.0","event_id":"sha256:b402e8d018701d34f228510bae3239faca723ffc1357fd6f34928d9a99f64f48"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:U6EDSYK3M5P3QWF4YUEGU3QMZ5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"PerturboLLaVA: Reducing Multimodal Hallucinations with Perturbative Visual Training","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Bo Zhang, Chenchen Jing, Chunhua Shen, Cong Chen, Fengyun Rao, Hao Chen, Mingyu Liu, Yizhou Zhou","submitted_at":"2025-03-09T07:07:03Z","abstract_excerpt":"This paper aims to address the challenge of hallucinations in Multimodal Large Language Models (MLLMs) particularly for dense image captioning tasks. To tackle the challenge, we identify the current lack of a metric that finely measures the caption quality in concept level. We hereby introduce HalFscore, a novel metric built upon the language graph and is designed to evaluate both the accuracy and completeness of dense captions at a granular level. Additionally, we identify the root cause of hallucination as the model's over-reliance on its language prior. To address this, we propose PerturboL"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.06486","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/2503.06486/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":"2026-07-05T10:27:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"vEpaw5J2ICfhSZ9EfW8bjOWlyPdGyHBUQiKCjPNzvRNBbrkQ3PqJSSkhjKhLX98khH9Nl580P88/3BFpJWOdDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-07T05:22:57.068613Z"},"content_sha256":"8e33663e60cefce0deb32e6c07ea990be2f9bf898bf7e134d2e064ea925c3749","schema_version":"1.0","event_id":"sha256:8e33663e60cefce0deb32e6c07ea990be2f9bf898bf7e134d2e064ea925c3749"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/bundle.json","state_url":"https://pith.science/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/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":"2026-08-07T05:22:57Z","links":{"resolver":"https://pith.science/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5","bundle":"https://pith.science/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/bundle.json","state":"https://pith.science/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/U6EDSYK3M5P3QWF4YUEGU3QMZ5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:U6EDSYK3M5P3QWF4YUEGU3QMZ5","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":"04540f22d8d0e341dfee01da002d7d2c279e8804a1b1df07d00f1c5c4e2c1da2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T07:07:03Z","title_canon_sha256":"4407fee008cbdb0ea314bfdef3767463c8de1c5ec367263381356634f5ee0b91"},"schema_version":"1.0","source":{"id":"2503.06486","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.06486","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"arxiv_version","alias_value":"2503.06486v1","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.06486","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_12","alias_value":"U6EDSYK3M5P3","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_16","alias_value":"U6EDSYK3M5P3QWF4","created_at":"2026-07-05T10:27:35Z"},{"alias_kind":"pith_short_8","alias_value":"U6EDSYK3","created_at":"2026-07-05T10:27:35Z"}],"graph_snapshots":[{"event_id":"sha256:8e33663e60cefce0deb32e6c07ea990be2f9bf898bf7e134d2e064ea925c3749","target":"graph","created_at":"2026-07-05T10:27:35Z","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/2503.06486/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"This paper aims to address the challenge of hallucinations in Multimodal Large Language Models (MLLMs) particularly for dense image captioning tasks. To tackle the challenge, we identify the current lack of a metric that finely measures the caption quality in concept level. We hereby introduce HalFscore, a novel metric built upon the language graph and is designed to evaluate both the accuracy and completeness of dense captions at a granular level. Additionally, we identify the root cause of hallucination as the model's over-reliance on its language prior. To address this, we propose PerturboL","authors_text":"Bo Zhang, Chenchen Jing, Chunhua Shen, Cong Chen, Fengyun Rao, Hao Chen, Mingyu Liu, Yizhou Zhou","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T07:07:03Z","title":"PerturboLLaVA: Reducing Multimodal Hallucinations with Perturbative Visual Training"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.06486","kind":"arxiv","version":1},"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:b402e8d018701d34f228510bae3239faca723ffc1357fd6f34928d9a99f64f48","target":"record","created_at":"2026-07-05T10:27:35Z","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":"04540f22d8d0e341dfee01da002d7d2c279e8804a1b1df07d00f1c5c4e2c1da2","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-09T07:07:03Z","title_canon_sha256":"4407fee008cbdb0ea314bfdef3767463c8de1c5ec367263381356634f5ee0b91"},"schema_version":"1.0","source":{"id":"2503.06486","kind":"arxiv","version":1}},"canonical_sha256":"a78839615b675fb858bcc5086a6e0ccf7059f10e99d8f355825e32bc520ebc16","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"a78839615b675fb858bcc5086a6e0ccf7059f10e99d8f355825e32bc520ebc16","first_computed_at":"2026-07-05T10:27:35.079188Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:27:35.079188Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"HELCyf8rDaWgllxmUxryrNNxMMVl8l/3h0iiD1g4KGjR8TFcr5A3kbWhvPBiHoWN0TtS3w58EyuXva+P3PScBg==","signature_status":"signed_v1","signed_at":"2026-07-05T10:27:35.079824Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.06486","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:b402e8d018701d34f228510bae3239faca723ffc1357fd6f34928d9a99f64f48","sha256:8e33663e60cefce0deb32e6c07ea990be2f9bf898bf7e134d2e064ea925c3749"],"state_sha256":"51ccf7ec7fe0eda18fbbfb12cfd2e5ab3ef229def869be568af7cb7d53e4b35f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"U8VbQaGI+JNdUMThpwJEbAvRo5CIiAl1sCF0hwkZkPd8v1HNAtn8IeMXE3RmUms/KZSCbDwdi9JNFWFHT1pCDQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-07T05:22:57.072917Z","bundle_sha256":"61c2861304389e8145f9f42680ca415ea9b2dfb6d93cf99094af4464feca4fe7"}}