{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:WQPC24LQMSJ3INWP4U6IC4ADHO","short_pith_number":"pith:WQPC24LQ","canonical_record":{"source":{"id":"2605.20035","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:55:16Z","cross_cats_sorted":[],"title_canon_sha256":"65ad8c0ba812f6fbabd6a76b6621e2298a6925a70de6a7a1b9345ed6d38b2ce1","abstract_canon_sha256":"d952448f510c17ba7b2ea7e536c8e7510383dcfa4856106396dc4482a28544c4"},"schema_version":"1.0"},"canonical_sha256":"b41e2d71706493b436cfe53c8170033b912a6ac95254d9714affa1de05b23e5a","source":{"kind":"arxiv","id":"2605.20035","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20035","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20035v1","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20035","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_12","alias_value":"WQPC24LQMSJ3","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_16","alias_value":"WQPC24LQMSJ3INWP","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_8","alias_value":"WQPC24LQ","created_at":"2026-05-20T02:05:59Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:WQPC24LQMSJ3INWP4U6IC4ADHO","target":"record","payload":{"canonical_record":{"source":{"id":"2605.20035","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:55:16Z","cross_cats_sorted":[],"title_canon_sha256":"65ad8c0ba812f6fbabd6a76b6621e2298a6925a70de6a7a1b9345ed6d38b2ce1","abstract_canon_sha256":"d952448f510c17ba7b2ea7e536c8e7510383dcfa4856106396dc4482a28544c4"},"schema_version":"1.0"},"canonical_sha256":"b41e2d71706493b436cfe53c8170033b912a6ac95254d9714affa1de05b23e5a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-20T02:05:59.716889Z","signature_b64":"u80niZX1zelQJVh75PuPFi1ZGTz7KKXrzmaZOsnQ/fLXvaQQO6n0HHPj+Y129e8v8jA3uDuV+AydOYPJbnFyAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"b41e2d71706493b436cfe53c8170033b912a6ac95254d9714affa1de05b23e5a","last_reissued_at":"2026-05-20T02:05:59.716446Z","signature_status":"signed_v1","first_computed_at":"2026-05-20T02:05:59.716446Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2605.20035","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":"2026-05-20T02:05:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"2UbkPdbeDIj+AZVt+VxcjQC5qHiN+K00KwYT0HaGFDzXZWrdAX2Wonz37X93bLoMYAuEPhBTNcYGepvHg9COBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T16:35:25.263360Z"},"content_sha256":"805ce0ab7edbb76a409f3fa9b0da5395744ecb01a9354821af7f2a7f92a7e240","schema_version":"1.0","event_id":"sha256:805ce0ab7edbb76a409f3fa9b0da5395744ecb01a9354821af7f2a7f92a7e240"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:WQPC24LQMSJ3INWP4U6IC4ADHO","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Stage-adaptive Token Selection for Efficient Omni-modal LLMs","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CV","authors_text":"Fengyun Rao, Jie Yang, Jing Lyu, Ruixiang Zhao, Tianyi Wang, Xirong Li, Zijie Xin","submitted_at":"2026-05-19T15:55:16Z","abstract_excerpt":"Omni-modal large language models (om-LLMs) achieve unified audio-visual understanding by encoding video and audio into temporally aligned token sequences interleaved at the window level. However, processing these dense non-textual tokens throughout the LLM incurs substantial computational overhead. Although training-free token selection can reduce this cost, existing methods either focus on visual-only inputs or prune om-LLM tokens only before the LLM with fixed per-modality ratios, failing to capture how cross-modal token importance evolves across layers. To address this limitation, we first "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20035","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/2605.20035/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":"2026-05-20T02:05:59Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IukViOanvkdazXVimfxuFGPdmh03SMXjUOBgaEWn4gnIJB7mC/JheJRGt3e06wa9atmAQlXZSzKFLXORoYpZDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-05-21T16:35:25.263764Z"},"content_sha256":"cabac73845245a05728a56cc346b13dd31e49cf276f9a2c5877a40634685e42d","schema_version":"1.0","event_id":"sha256:cabac73845245a05728a56cc346b13dd31e49cf276f9a2c5877a40634685e42d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/bundle.json","state_url":"https://pith.science/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/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":"2026-05-21T16:35:25Z","links":{"resolver":"https://pith.science/pith/WQPC24LQMSJ3INWP4U6IC4ADHO","bundle":"https://pith.science/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/bundle.json","state":"https://pith.science/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/state.json","well_known_bundle":"https://pith.science/.well-known/pith/WQPC24LQMSJ3INWP4U6IC4ADHO/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:WQPC24LQMSJ3INWP4U6IC4ADHO","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":"d952448f510c17ba7b2ea7e536c8e7510383dcfa4856106396dc4482a28544c4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:55:16Z","title_canon_sha256":"65ad8c0ba812f6fbabd6a76b6621e2298a6925a70de6a7a1b9345ed6d38b2ce1"},"schema_version":"1.0","source":{"id":"2605.20035","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2605.20035","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"arxiv_version","alias_value":"2605.20035v1","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2605.20035","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_12","alias_value":"WQPC24LQMSJ3","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_16","alias_value":"WQPC24LQMSJ3INWP","created_at":"2026-05-20T02:05:59Z"},{"alias_kind":"pith_short_8","alias_value":"WQPC24LQ","created_at":"2026-05-20T02:05:59Z"}],"graph_snapshots":[{"event_id":"sha256:cabac73845245a05728a56cc346b13dd31e49cf276f9a2c5877a40634685e42d","target":"graph","created_at":"2026-05-20T02:05:59Z","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/2605.20035/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Omni-modal large language models (om-LLMs) achieve unified audio-visual understanding by encoding video and audio into temporally aligned token sequences interleaved at the window level. However, processing these dense non-textual tokens throughout the LLM incurs substantial computational overhead. Although training-free token selection can reduce this cost, existing methods either focus on visual-only inputs or prune om-LLM tokens only before the LLM with fixed per-modality ratios, failing to capture how cross-modal token importance evolves across layers. To address this limitation, we first ","authors_text":"Fengyun Rao, Jie Yang, Jing Lyu, Ruixiang Zhao, Tianyi Wang, Xirong Li, Zijie Xin","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:55:16Z","title":"Stage-adaptive Token Selection for Efficient Omni-modal LLMs"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2605.20035","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":"sha256:805ce0ab7edbb76a409f3fa9b0da5395744ecb01a9354821af7f2a7f92a7e240","target":"record","created_at":"2026-05-20T02:05:59Z","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":"d952448f510c17ba7b2ea7e536c8e7510383dcfa4856106396dc4482a28544c4","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CV","submitted_at":"2026-05-19T15:55:16Z","title_canon_sha256":"65ad8c0ba812f6fbabd6a76b6621e2298a6925a70de6a7a1b9345ed6d38b2ce1"},"schema_version":"1.0","source":{"id":"2605.20035","kind":"arxiv","version":1}},"canonical_sha256":"b41e2d71706493b436cfe53c8170033b912a6ac95254d9714affa1de05b23e5a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"b41e2d71706493b436cfe53c8170033b912a6ac95254d9714affa1de05b23e5a","first_computed_at":"2026-05-20T02:05:59.716446Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-20T02:05:59.716446Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"u80niZX1zelQJVh75PuPFi1ZGTz7KKXrzmaZOsnQ/fLXvaQQO6n0HHPj+Y129e8v8jA3uDuV+AydOYPJbnFyAA==","signature_status":"signed_v1","signed_at":"2026-05-20T02:05:59.716889Z","signed_message":"canonical_sha256_bytes"},"source_id":"2605.20035","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:805ce0ab7edbb76a409f3fa9b0da5395744ecb01a9354821af7f2a7f92a7e240","sha256:cabac73845245a05728a56cc346b13dd31e49cf276f9a2c5877a40634685e42d"],"state_sha256":"1e92fc1afd36f3b4129a8fd826dde459375b485430b9f31ca5c8bc265f8d6e53"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"YYOG10xmoGhP/tEOR+EZRZD+kk1ezYqU1EF/+VhagBza4d99Cb7ijize2EKSylqLYcMZwjsK9HQXwlpGXgnrCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-05-21T16:35:25.266029Z","bundle_sha256":"c5f51c3b22416ad27c48ae59eb9db109bf9c819fdfdfee4279ab45884a79220b"}}