{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:MTOQ2BKG6PN622NNNBAAWQ6QKA","short_pith_number":"pith:MTOQ2BKG","canonical_record":{"source":{"id":"2503.20540","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-26T13:38:10Z","cross_cats_sorted":[],"title_canon_sha256":"61f9e4cac6702adefb01deb7d6efc62a70e486fc4bf6541a3162575b1e1a5331","abstract_canon_sha256":"9aca959c21c6d72d460e9515b55fba1a1bbceec2f80509030434e5dd365e1c2c"},"schema_version":"1.0"},"canonical_sha256":"64dd0d0546f3dbed69ad68400b43d050083da2124221ab9241138ec4cf63b1ef","source":{"kind":"arxiv","id":"2503.20540","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.20540","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"arxiv_version","alias_value":"2503.20540v1","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.20540","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_12","alias_value":"MTOQ2BKG6PN6","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_16","alias_value":"MTOQ2BKG6PN622NN","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_8","alias_value":"MTOQ2BKG","created_at":"2026-07-05T10:39:39Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:MTOQ2BKG6PN622NNNBAAWQ6QKA","target":"record","payload":{"canonical_record":{"source":{"id":"2503.20540","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-26T13:38:10Z","cross_cats_sorted":[],"title_canon_sha256":"61f9e4cac6702adefb01deb7d6efc62a70e486fc4bf6541a3162575b1e1a5331","abstract_canon_sha256":"9aca959c21c6d72d460e9515b55fba1a1bbceec2f80509030434e5dd365e1c2c"},"schema_version":"1.0"},"canonical_sha256":"64dd0d0546f3dbed69ad68400b43d050083da2124221ab9241138ec4cf63b1ef","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:39:39.873151Z","signature_b64":"LcZW2jJ1iUoHZAPqf76TChkxurFM3cCT2S85u1SlpG8t6HE6YlQJg56+AMd2R/qRetCXzTg9k0N1QUJfGZASCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"64dd0d0546f3dbed69ad68400b43d050083da2124221ab9241138ec4cf63b1ef","last_reissued_at":"2026-07-05T10:39:39.872622Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:39:39.872622Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2503.20540","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":"2026-07-05T10:39:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"OwjPLzj67FDnaMHtMZbvMLWPip5/uWU4wRs3J4PL5pz7/BtXCgJNtAcrqQQ2VNlb0ONijAA4CfbmiPn+xVgpCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T08:54:46.395909Z"},"content_sha256":"4ac50ab2bd61313b301f50c8638a7ed7ca80424b5662ea88064949bc9af73a57","schema_version":"1.0","event_id":"sha256:4ac50ab2bd61313b301f50c8638a7ed7ca80424b5662ea88064949bc9af73a57"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:MTOQ2BKG6PN622NNNBAAWQ6QKA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Beyond Intermediate States: Explaining Visual Redundancy through Language","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Anran Zhang, Bowen Cao, Dingchen Yang, Guang Chen, Weibo Gu, Winston Hu","submitted_at":"2025-03-26T13:38:10Z","abstract_excerpt":"Multi-modal Large Langue Models (MLLMs) often process thousands of visual tokens, which consume a significant portion of the context window and impose a substantial computational burden. Prior work has empirically explored visual token pruning methods based on MLLMs' intermediate states (e.g., attention scores). However, they have limitations in precisely defining visual redundancy due to their inability to capture the influence of visual tokens on MLLMs' visual understanding (i.e., the predicted probabilities for textual token candidates). To address this issue, we manipulate the visual input"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.20540","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/2503.20540/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":"2026-07-05T10:39:39Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"H6TS4ikHocSzA++UD8AVTZwwEyyJApriDrwbZmfdlwDvAkQ7lA1BURwsTRtxEQqi0tgiqxmy9ImAfn6L4ZLlDQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-04T08:54:46.396423Z"},"content_sha256":"2e9c8c907705a51ae4630610eeff9ba60f34fd364a2a6ed6f3e38d3e1ef7c910","schema_version":"1.0","event_id":"sha256:2e9c8c907705a51ae4630610eeff9ba60f34fd364a2a6ed6f3e38d3e1ef7c910"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/bundle.json","state_url":"https://pith.science/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/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":"2026-08-04T08:54:46Z","links":{"resolver":"https://pith.science/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA","bundle":"https://pith.science/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/bundle.json","state":"https://pith.science/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MTOQ2BKG6PN622NNNBAAWQ6QKA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:MTOQ2BKG6PN622NNNBAAWQ6QKA","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":"9aca959c21c6d72d460e9515b55fba1a1bbceec2f80509030434e5dd365e1c2c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-26T13:38:10Z","title_canon_sha256":"61f9e4cac6702adefb01deb7d6efc62a70e486fc4bf6541a3162575b1e1a5331"},"schema_version":"1.0","source":{"id":"2503.20540","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2503.20540","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"arxiv_version","alias_value":"2503.20540v1","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2503.20540","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_12","alias_value":"MTOQ2BKG6PN6","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_16","alias_value":"MTOQ2BKG6PN622NN","created_at":"2026-07-05T10:39:39Z"},{"alias_kind":"pith_short_8","alias_value":"MTOQ2BKG","created_at":"2026-07-05T10:39:39Z"}],"graph_snapshots":[{"event_id":"sha256:2e9c8c907705a51ae4630610eeff9ba60f34fd364a2a6ed6f3e38d3e1ef7c910","target":"graph","created_at":"2026-07-05T10:39:39Z","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/2503.20540/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multi-modal Large Langue Models (MLLMs) often process thousands of visual tokens, which consume a significant portion of the context window and impose a substantial computational burden. Prior work has empirically explored visual token pruning methods based on MLLMs' intermediate states (e.g., attention scores). However, they have limitations in precisely defining visual redundancy due to their inability to capture the influence of visual tokens on MLLMs' visual understanding (i.e., the predicted probabilities for textual token candidates). To address this issue, we manipulate the visual input","authors_text":"Anran Zhang, Bowen Cao, Dingchen Yang, Guang Chen, Weibo Gu, Winston Hu","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-26T13:38:10Z","title":"Beyond Intermediate States: Explaining Visual Redundancy through Language"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2503.20540","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":"sha256:4ac50ab2bd61313b301f50c8638a7ed7ca80424b5662ea88064949bc9af73a57","target":"record","created_at":"2026-07-05T10:39:39Z","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":"9aca959c21c6d72d460e9515b55fba1a1bbceec2f80509030434e5dd365e1c2c","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CV","submitted_at":"2025-03-26T13:38:10Z","title_canon_sha256":"61f9e4cac6702adefb01deb7d6efc62a70e486fc4bf6541a3162575b1e1a5331"},"schema_version":"1.0","source":{"id":"2503.20540","kind":"arxiv","version":1}},"canonical_sha256":"64dd0d0546f3dbed69ad68400b43d050083da2124221ab9241138ec4cf63b1ef","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"64dd0d0546f3dbed69ad68400b43d050083da2124221ab9241138ec4cf63b1ef","first_computed_at":"2026-07-05T10:39:39.872622Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:39:39.872622Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"LcZW2jJ1iUoHZAPqf76TChkxurFM3cCT2S85u1SlpG8t6HE6YlQJg56+AMd2R/qRetCXzTg9k0N1QUJfGZASCQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:39:39.873151Z","signed_message":"canonical_sha256_bytes"},"source_id":"2503.20540","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:4ac50ab2bd61313b301f50c8638a7ed7ca80424b5662ea88064949bc9af73a57","sha256:2e9c8c907705a51ae4630610eeff9ba60f34fd364a2a6ed6f3e38d3e1ef7c910"],"state_sha256":"53438e02b393b2b9bc7f8f3e11ff64891dd471c96d32263648f9a0a1b7d64745"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"edlwP1el8bgib4xNYeh0gcet7ip56s7gMqDQ8kL+HceBTlNsRCPhqhpr7AkncuLLa9TOs4ACD7xGvGffTNuGCQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-04T08:54:46.400985Z","bundle_sha256":"77b88776324d1682c2f1e3a8721f3ecf7defa03193441c9e693a3ce2249d977d"}}