{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:MCO75XINF6PKYBANZ3DET5GBK4","short_pith_number":"pith:MCO75XIN","canonical_record":{"source":{"id":"2606.27161","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-25T15:29:37Z","cross_cats_sorted":[],"title_canon_sha256":"7c07a8fde486c9f3f20a1800d41f707e42b97c0f6e1132dcf9c82f8a19635c13","abstract_canon_sha256":"8c276f88865b6ef4daa2132422d95640423675a0f471e785417e12e1398ac448"},"schema_version":"1.0"},"canonical_sha256":"609dfedd0d2f9eac040dcec649f4c157092f7bd5c5430246ad2f47473ba4c397","source":{"kind":"arxiv","id":"2606.27161","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.27161","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"arxiv_version","alias_value":"2606.27161v1","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27161","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_12","alias_value":"MCO75XINF6PK","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_16","alias_value":"MCO75XINF6PKYBAN","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_8","alias_value":"MCO75XIN","created_at":"2026-06-26T01:16:11Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:MCO75XINF6PKYBANZ3DET5GBK4","target":"record","payload":{"canonical_record":{"source":{"id":"2606.27161","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-25T15:29:37Z","cross_cats_sorted":[],"title_canon_sha256":"7c07a8fde486c9f3f20a1800d41f707e42b97c0f6e1132dcf9c82f8a19635c13","abstract_canon_sha256":"8c276f88865b6ef4daa2132422d95640423675a0f471e785417e12e1398ac448"},"schema_version":"1.0"},"canonical_sha256":"609dfedd0d2f9eac040dcec649f4c157092f7bd5c5430246ad2f47473ba4c397","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-06-26T01:16:11.903926Z","signature_b64":"z5lAXpe8jK+5gBD/NBINQE7rXSRCeNyhlfH9r8oVG7NC02bqxaI6yDXaF6LYoVQKL975CHIhuglg1TPMslziDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"609dfedd0d2f9eac040dcec649f4c157092f7bd5c5430246ad2f47473ba4c397","last_reissued_at":"2026-06-26T01:16:11.903571Z","signature_status":"signed_v1","first_computed_at":"2026-06-26T01:16:11.903571Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2606.27161","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":"2026-06-26T01:16:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"7dBa3gnGd6WBWAtnREAHMhgNgjS7CMM40WvM1mJxldl2JtcEHguEpLvYINC4agZG9CG2qBhmaMpbCjQGIvIKDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T23:46:19.504758Z"},"content_sha256":"3eb94ea2afb4225e459fc99a67f805b891aabb3be1457fe18f121472b9a92158","schema_version":"1.0","event_id":"sha256:3eb94ea2afb4225e459fc99a67f805b891aabb3be1457fe18f121472b9a92158"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:MCO75XINF6PKYBANZ3DET5GBK4","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"TOPS: First-Principles Visual Token Pruning via Constructing Token Optimal Preservation Sets for Efficient MLLM Inference","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Chenxi Li, Guangyan Gan, Jiajun Cao, Lin William Cong, Qizhe Zhang, Rui Huang, Shanghang Zhang, Tinghao Wang, Wenya Wang, Yaosong Du, Yichen Guo, Yuan Zhang, Zheng Lu, Zhirong Shen","submitted_at":"2026-06-25T15:29:37Z","abstract_excerpt":"Multimodal large language models (MLLMs) have achieved strong multimodal reasoning capabilities, but their efficiency is limited by the large number of visual tokens, which introduces substantial computational overhead. Visual token pruning offers a natural solution, yet existing methods are imperfect: attention-based criteria tend to retain redundant tokens, while diversity-based criteria are often agnostic to user instructions. Even methods that combine multiple criteria still lack a principled formulation of the intrinsic objective of token pruning. In this paper, we revisit visual token pr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27161","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/2606.27161/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":"2026-06-26T01:16:11Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"cRZyDK7DF/9hKvZaPLVW3Fi0xN8caLe6f5IIhdtSXss+dtVby/K1h0O2OJDNJ9uLHfNs+LxbIa/hvtEs0k9kAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-04T23:46:19.505126Z"},"content_sha256":"12c91d180e3160c535f9d42285b0c3ce413beccee03f483999524423681ba1ca","schema_version":"1.0","event_id":"sha256:12c91d180e3160c535f9d42285b0c3ce413beccee03f483999524423681ba1ca"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/MCO75XINF6PKYBANZ3DET5GBK4/bundle.json","state_url":"https://pith.science/pith/MCO75XINF6PKYBANZ3DET5GBK4/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/MCO75XINF6PKYBANZ3DET5GBK4/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":"2026-07-04T23:46:19Z","links":{"resolver":"https://pith.science/pith/MCO75XINF6PKYBANZ3DET5GBK4","bundle":"https://pith.science/pith/MCO75XINF6PKYBANZ3DET5GBK4/bundle.json","state":"https://pith.science/pith/MCO75XINF6PKYBANZ3DET5GBK4/state.json","well_known_bundle":"https://pith.science/.well-known/pith/MCO75XINF6PKYBANZ3DET5GBK4/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:MCO75XINF6PKYBANZ3DET5GBK4","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":"8c276f88865b6ef4daa2132422d95640423675a0f471e785417e12e1398ac448","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-25T15:29:37Z","title_canon_sha256":"7c07a8fde486c9f3f20a1800d41f707e42b97c0f6e1132dcf9c82f8a19635c13"},"schema_version":"1.0","source":{"id":"2606.27161","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2606.27161","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"arxiv_version","alias_value":"2606.27161v1","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2606.27161","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_12","alias_value":"MCO75XINF6PK","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_16","alias_value":"MCO75XINF6PKYBAN","created_at":"2026-06-26T01:16:11Z"},{"alias_kind":"pith_short_8","alias_value":"MCO75XIN","created_at":"2026-06-26T01:16:11Z"}],"graph_snapshots":[{"event_id":"sha256:12c91d180e3160c535f9d42285b0c3ce413beccee03f483999524423681ba1ca","target":"graph","created_at":"2026-06-26T01:16:11Z","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/2606.27161/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Multimodal large language models (MLLMs) have achieved strong multimodal reasoning capabilities, but their efficiency is limited by the large number of visual tokens, which introduces substantial computational overhead. Visual token pruning offers a natural solution, yet existing methods are imperfect: attention-based criteria tend to retain redundant tokens, while diversity-based criteria are often agnostic to user instructions. Even methods that combine multiple criteria still lack a principled formulation of the intrinsic objective of token pruning. In this paper, we revisit visual token pr","authors_text":"Chenxi Li, Guangyan Gan, Jiajun Cao, Lin William Cong, Qizhe Zhang, Rui Huang, Shanghang Zhang, Tinghao Wang, Wenya Wang, Yaosong Du, Yichen Guo, Yuan Zhang, Zheng Lu, Zhirong Shen","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-25T15:29:37Z","title":"TOPS: First-Principles Visual Token Pruning via Constructing Token Optimal Preservation Sets for Efficient MLLM Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2606.27161","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":"sha256:3eb94ea2afb4225e459fc99a67f805b891aabb3be1457fe18f121472b9a92158","target":"record","created_at":"2026-06-26T01:16:11Z","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":"8c276f88865b6ef4daa2132422d95640423675a0f471e785417e12e1398ac448","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-06-25T15:29:37Z","title_canon_sha256":"7c07a8fde486c9f3f20a1800d41f707e42b97c0f6e1132dcf9c82f8a19635c13"},"schema_version":"1.0","source":{"id":"2606.27161","kind":"arxiv","version":1}},"canonical_sha256":"609dfedd0d2f9eac040dcec649f4c157092f7bd5c5430246ad2f47473ba4c397","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"609dfedd0d2f9eac040dcec649f4c157092f7bd5c5430246ad2f47473ba4c397","first_computed_at":"2026-06-26T01:16:11.903571Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-06-26T01:16:11.903571Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"z5lAXpe8jK+5gBD/NBINQE7rXSRCeNyhlfH9r8oVG7NC02bqxaI6yDXaF6LYoVQKL975CHIhuglg1TPMslziDQ==","signature_status":"signed_v1","signed_at":"2026-06-26T01:16:11.903926Z","signed_message":"canonical_sha256_bytes"},"source_id":"2606.27161","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:3eb94ea2afb4225e459fc99a67f805b891aabb3be1457fe18f121472b9a92158","sha256:12c91d180e3160c535f9d42285b0c3ce413beccee03f483999524423681ba1ca"],"state_sha256":"87534ef71536bc06783657597a2bfa03c9677e28640deb335ebf9feb6bff2a96"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"48nz0W/Sel1Pp6nASRuly/cAWJQJ4GEPtJIZrOmy3zdVVMJKKHTS17to3PPHml6fDJgms33/jixRQUMj+SvIAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-04T23:46:19.507132Z","bundle_sha256":"037ac48bac7655c51086d0f3d306e719380cc7f56ea18e0e753b51dce4949f75"}}