{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:T6ZZAANWBAGFIXS75NIWLHTR6H","short_pith_number":"pith:T6ZZAANW","canonical_record":{"source":{"id":"2405.14612","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T14:24:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bd334ec9eadda919047c3a8c14fb49a637712c631048d15e7d5f410004cecd5d","abstract_canon_sha256":"c9494811007294ffc2eb92a242eb8e43e1f6ad315c48b15ba2945bce6db62d57"},"schema_version":"1.0"},"canonical_sha256":"9fb39001b6080c545e5feb51659e71f1e2d458b4dd35c0be6800f5965475643e","source":{"kind":"arxiv","id":"2405.14612","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.14612","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"arxiv_version","alias_value":"2405.14612v2","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14612","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_12","alias_value":"T6ZZAANWBAGF","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_16","alias_value":"T6ZZAANWBAGFIXS7","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_8","alias_value":"T6ZZAANW","created_at":"2026-07-05T08:24:06Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:T6ZZAANWBAGFIXS75NIWLHTR6H","target":"record","payload":{"canonical_record":{"source":{"id":"2405.14612","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T14:24:23Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"bd334ec9eadda919047c3a8c14fb49a637712c631048d15e7d5f410004cecd5d","abstract_canon_sha256":"c9494811007294ffc2eb92a242eb8e43e1f6ad315c48b15ba2945bce6db62d57"},"schema_version":"1.0"},"canonical_sha256":"9fb39001b6080c545e5feb51659e71f1e2d458b4dd35c0be6800f5965475643e","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:24:06.305326Z","signature_b64":"j7N+nd18XWNSLQSDN8gIj7iPrLFSF167J4W0pR9ut3Byc+A1yVo98IbepMljkSd7FBZL+Ntm7to39yW87/kWBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9fb39001b6080c545e5feb51659e71f1e2d458b4dd35c0be6800f5965475643e","last_reissued_at":"2026-07-05T08:24:06.304839Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:24:06.304839Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2405.14612","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:24:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"inOYyq+4gXSj6+qPgd5ysGWgsct5f243S/ofZYQTssx2zOkW8AEjTB20w3zmxIcS27R+UGarAB7J/q7wDF+CDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:57:06.153298Z"},"content_sha256":"797501e69f3a4c1958c4802c9532015708e00b4710fab4f578122fa7437059a0","schema_version":"1.0","event_id":"sha256:797501e69f3a4c1958c4802c9532015708e00b4710fab4f578122fa7437059a0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:T6ZZAANWBAGFIXS75NIWLHTR6H","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Explaining Multi-modal Large Language Models by Analyzing their Vision Perception","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CV","authors_text":"Giacomo Boracchi, Loris Giulivi","submitted_at":"2024-05-23T14:24:23Z","abstract_excerpt":"Multi-modal Large Language Models (MLLMs) have demonstrated remarkable capabilities in understanding and generating content across various modalities, such as images and text. However, their interpretability remains a challenge, hindering their adoption in critical applications. This research proposes a novel approach to enhance the interpretability of MLLMs by focusing on the image embedding component. We combine an open-world localization model with a MLLM, thus creating a new architecture able to simultaneously produce text and object localization outputs from the same vision embedding. The"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14612","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/2405.14612/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":"2026-07-05T08:24:06Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9vtrsmYriZQ/ZY3PFlMacHH5uJwugI42U6hP/cePdixaMlTfM1J+Nk4xkyV07n/LdAo8qO26aQg1fAcZR1lMAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T20:57:06.153804Z"},"content_sha256":"b89ff6058f163040c64a7d7c3229e9c44764e5e3534fd4a132fdce5a883be283","schema_version":"1.0","event_id":"sha256:b89ff6058f163040c64a7d7c3229e9c44764e5e3534fd4a132fdce5a883be283"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/bundle.json","state_url":"https://pith.science/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/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":"2026-08-09T20:57:06Z","links":{"resolver":"https://pith.science/pith/T6ZZAANWBAGFIXS75NIWLHTR6H","bundle":"https://pith.science/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/bundle.json","state":"https://pith.science/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/state.json","well_known_bundle":"https://pith.science/.well-known/pith/T6ZZAANWBAGFIXS75NIWLHTR6H/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:T6ZZAANWBAGFIXS75NIWLHTR6H","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":"c9494811007294ffc2eb92a242eb8e43e1f6ad315c48b15ba2945bce6db62d57","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T14:24:23Z","title_canon_sha256":"bd334ec9eadda919047c3a8c14fb49a637712c631048d15e7d5f410004cecd5d"},"schema_version":"1.0","source":{"id":"2405.14612","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2405.14612","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"arxiv_version","alias_value":"2405.14612v2","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.14612","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_12","alias_value":"T6ZZAANWBAGF","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_16","alias_value":"T6ZZAANWBAGFIXS7","created_at":"2026-07-05T08:24:06Z"},{"alias_kind":"pith_short_8","alias_value":"T6ZZAANW","created_at":"2026-07-05T08:24:06Z"}],"graph_snapshots":[{"event_id":"sha256:b89ff6058f163040c64a7d7c3229e9c44764e5e3534fd4a132fdce5a883be283","target":"graph","created_at":"2026-07-05T08:24:06Z","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/2405.14612/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multi-modal Large Language Models (MLLMs) have demonstrated remarkable capabilities in understanding and generating content across various modalities, such as images and text. However, their interpretability remains a challenge, hindering their adoption in critical applications. This research proposes a novel approach to enhance the interpretability of MLLMs by focusing on the image embedding component. We combine an open-world localization model with a MLLM, thus creating a new architecture able to simultaneously produce text and object localization outputs from the same vision embedding. The","authors_text":"Giacomo Boracchi, Loris Giulivi","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T14:24:23Z","title":"Explaining Multi-modal Large Language Models by Analyzing their Vision Perception"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2405.14612","kind":"arxiv","version":2},"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":"sha256:797501e69f3a4c1958c4802c9532015708e00b4710fab4f578122fa7437059a0","target":"record","created_at":"2026-07-05T08:24:06Z","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":"c9494811007294ffc2eb92a242eb8e43e1f6ad315c48b15ba2945bce6db62d57","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.CV","submitted_at":"2024-05-23T14:24:23Z","title_canon_sha256":"bd334ec9eadda919047c3a8c14fb49a637712c631048d15e7d5f410004cecd5d"},"schema_version":"1.0","source":{"id":"2405.14612","kind":"arxiv","version":2}},"canonical_sha256":"9fb39001b6080c545e5feb51659e71f1e2d458b4dd35c0be6800f5965475643e","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9fb39001b6080c545e5feb51659e71f1e2d458b4dd35c0be6800f5965475643e","first_computed_at":"2026-07-05T08:24:06.304839Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:24:06.304839Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"j7N+nd18XWNSLQSDN8gIj7iPrLFSF167J4W0pR9ut3Byc+A1yVo98IbepMljkSd7FBZL+Ntm7to39yW87/kWBg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:24:06.305326Z","signed_message":"canonical_sha256_bytes"},"source_id":"2405.14612","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:797501e69f3a4c1958c4802c9532015708e00b4710fab4f578122fa7437059a0","sha256:b89ff6058f163040c64a7d7c3229e9c44764e5e3534fd4a132fdce5a883be283"],"state_sha256":"a149865a60587eeb2cdfc9fae1f065f82001d058886f9163e7b0df661c17f8b4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mjBonQ3fh5mvzLcmfbuqOtlVN0UkHWvaQkrYWA+ojskEyOHYk0cT5VWRpaoveqQHdCwMSar0cyBBB1dVdcomBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T20:57:06.208091Z","bundle_sha256":"8460da71ad266e47c4a1a3607da464246e0d123d07e5e1edb5a08242df019dc8"}}