{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:SQOCCVGED5I5AKDWEGMIBMX3RC","short_pith_number":"pith:SQOCCVGE","canonical_record":{"source":{"id":"2306.14393","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-26T03:06:57Z","cross_cats_sorted":[],"title_canon_sha256":"949270cafa72954217729e8b7fdcabefbe09e18ed198561a1ff6eea316c88e72","abstract_canon_sha256":"ce045da0498b18ef3364913aeb74ea5479f7eef8bba320314275ebc02d00dd35"},"schema_version":"1.0"},"canonical_sha256":"941c2154c41f51d02876219880b2fb88864e08ea88b643d92f4c21325ef8bee9","source":{"kind":"arxiv","id":"2306.14393","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.14393","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"arxiv_version","alias_value":"2306.14393v1","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.14393","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_12","alias_value":"SQOCCVGED5I5","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_16","alias_value":"SQOCCVGED5I5AKDW","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_8","alias_value":"SQOCCVGE","created_at":"2026-07-05T06:24:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:SQOCCVGED5I5AKDWEGMIBMX3RC","target":"record","payload":{"canonical_record":{"source":{"id":"2306.14393","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-26T03:06:57Z","cross_cats_sorted":[],"title_canon_sha256":"949270cafa72954217729e8b7fdcabefbe09e18ed198561a1ff6eea316c88e72","abstract_canon_sha256":"ce045da0498b18ef3364913aeb74ea5479f7eef8bba320314275ebc02d00dd35"},"schema_version":"1.0"},"canonical_sha256":"941c2154c41f51d02876219880b2fb88864e08ea88b643d92f4c21325ef8bee9","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:24:35.976610Z","signature_b64":"On3ZBXqOL5faemdUXTwTbvC8J+0KlXDbkwaj3EenVXXWRSdGtVRr5jyByQ05mE7LBnU7TTO8U85e5sc1jaQSCA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"941c2154c41f51d02876219880b2fb88864e08ea88b643d92f4c21325ef8bee9","last_reissued_at":"2026-07-05T06:24:35.976172Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:24:35.976172Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2306.14393","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":"2026-07-05T06:24:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"utW4ZWG3HySqAF7JO4sdftQDlSvsNwNdHXBVB0ju1qAx2rinnLZMCJyiD4x1uxHx5Q7oXdxn1OpDJ29zxcm8CA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T23:14:20.515879Z"},"content_sha256":"9fbe9c4da3c36b65755838550c0de6af51e9aded24507852b007aa91df610cac","schema_version":"1.0","event_id":"sha256:9fbe9c4da3c36b65755838550c0de6af51e9aded24507852b007aa91df610cac"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:SQOCCVGED5I5AKDWEGMIBMX3RC","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Hao Sun, Jiahang Xu, Junyan Li, Li Lyna Zhang, Mao Yang, Qi Zhang, Shaoguang Yan, Ting Cao, Weiwei Deng, Yujing Wang, Yunqing Xia, Yuqing Yang","submitted_at":"2023-06-26T03:06:57Z","abstract_excerpt":"Deploying pre-trained transformer models like BERT on downstream tasks in resource-constrained scenarios is challenging due to their high inference cost, which grows rapidly with input sequence length. In this work, we propose a constraint-aware and ranking-distilled token pruning method ToP, which selectively removes unnecessary tokens as input sequence passes through layers, allowing the model to improve online inference speed while preserving accuracy. ToP overcomes the limitation of inaccurate token importance ranking in the conventional self-attention mechanism through a ranking-distilled"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.14393","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/2306.14393/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":"2026-07-05T06:24:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"9SXYBsEtd9/BAs6EhUloYxG+SddrjoJQssOMJpmf4/TwxBjI2aOx/cRjZDxI7O0GMoruuO7Eo9rp5gHfQQXbDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-05T23:14:20.516500Z"},"content_sha256":"60fc2c2037c609cea79b0fb53fc2bd783ab3dc2546da8e5492983f3247a4350e","schema_version":"1.0","event_id":"sha256:60fc2c2037c609cea79b0fb53fc2bd783ab3dc2546da8e5492983f3247a4350e"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/bundle.json","state_url":"https://pith.science/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/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":"2026-08-05T23:14:20Z","links":{"resolver":"https://pith.science/pith/SQOCCVGED5I5AKDWEGMIBMX3RC","bundle":"https://pith.science/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/bundle.json","state":"https://pith.science/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SQOCCVGED5I5AKDWEGMIBMX3RC/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:SQOCCVGED5I5AKDWEGMIBMX3RC","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":"ce045da0498b18ef3364913aeb74ea5479f7eef8bba320314275ebc02d00dd35","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-26T03:06:57Z","title_canon_sha256":"949270cafa72954217729e8b7fdcabefbe09e18ed198561a1ff6eea316c88e72"},"schema_version":"1.0","source":{"id":"2306.14393","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2306.14393","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"arxiv_version","alias_value":"2306.14393v1","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2306.14393","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_12","alias_value":"SQOCCVGED5I5","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_16","alias_value":"SQOCCVGED5I5AKDW","created_at":"2026-07-05T06:24:35Z"},{"alias_kind":"pith_short_8","alias_value":"SQOCCVGE","created_at":"2026-07-05T06:24:35Z"}],"graph_snapshots":[{"event_id":"sha256:60fc2c2037c609cea79b0fb53fc2bd783ab3dc2546da8e5492983f3247a4350e","target":"graph","created_at":"2026-07-05T06:24:35Z","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/2306.14393/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Deploying pre-trained transformer models like BERT on downstream tasks in resource-constrained scenarios is challenging due to their high inference cost, which grows rapidly with input sequence length. In this work, we propose a constraint-aware and ranking-distilled token pruning method ToP, which selectively removes unnecessary tokens as input sequence passes through layers, allowing the model to improve online inference speed while preserving accuracy. ToP overcomes the limitation of inaccurate token importance ranking in the conventional self-attention mechanism through a ranking-distilled","authors_text":"Hao Sun, Jiahang Xu, Junyan Li, Li Lyna Zhang, Mao Yang, Qi Zhang, Shaoguang Yan, Ting Cao, Weiwei Deng, Yujing Wang, Yunqing Xia, Yuqing Yang","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-26T03:06:57Z","title":"Constraint-aware and Ranking-distilled Token Pruning for Efficient Transformer Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2306.14393","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":"sha256:9fbe9c4da3c36b65755838550c0de6af51e9aded24507852b007aa91df610cac","target":"record","created_at":"2026-07-05T06:24:35Z","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":"ce045da0498b18ef3364913aeb74ea5479f7eef8bba320314275ebc02d00dd35","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.CL","submitted_at":"2023-06-26T03:06:57Z","title_canon_sha256":"949270cafa72954217729e8b7fdcabefbe09e18ed198561a1ff6eea316c88e72"},"schema_version":"1.0","source":{"id":"2306.14393","kind":"arxiv","version":1}},"canonical_sha256":"941c2154c41f51d02876219880b2fb88864e08ea88b643d92f4c21325ef8bee9","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"941c2154c41f51d02876219880b2fb88864e08ea88b643d92f4c21325ef8bee9","first_computed_at":"2026-07-05T06:24:35.976172Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T06:24:35.976172Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"On3ZBXqOL5faemdUXTwTbvC8J+0KlXDbkwaj3EenVXXWRSdGtVRr5jyByQ05mE7LBnU7TTO8U85e5sc1jaQSCA==","signature_status":"signed_v1","signed_at":"2026-07-05T06:24:35.976610Z","signed_message":"canonical_sha256_bytes"},"source_id":"2306.14393","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:9fbe9c4da3c36b65755838550c0de6af51e9aded24507852b007aa91df610cac","sha256:60fc2c2037c609cea79b0fb53fc2bd783ab3dc2546da8e5492983f3247a4350e"],"state_sha256":"e161fc9dbed3881ddd4a43085d3fd250ec23517746672fbcf583fa400fe4a1cd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"u0PQqv+Iseug0NW6wSJzG6o5CvOttsS4RPYjF1hJBKsK94Nb+98tPa4OT/mH9vZCE0AqWic/n99wzyz13yFOCA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-05T23:14:20.521872Z","bundle_sha256":"2a1ee4993ab3170f3c658d7280ebdbc830705e38e0212cc11b499e8111405e31"}}