{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2020:VSGBFURZPB65MFCRCIIRAN3EZA","short_pith_number":"pith:VSGBFURZ","canonical_record":{"source":{"id":"2004.12993","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-27T17:58:05Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1bbd1d03b192a4255bb12439d73db6d9410c00fcd3e07597e287abf41e934e8e","abstract_canon_sha256":"23e10bbf062d188fdb9f6ad92788bb2d3ddcc1e69ccae94a6fba1ca60af2b324"},"schema_version":"1.0"},"canonical_sha256":"ac8c12d239787dd614511211103764c827c80c977ff83cf689833bfe586fb6c8","source":{"kind":"arxiv","id":"2004.12993","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.12993","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"arxiv_version","alias_value":"2004.12993v1","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.12993","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_12","alias_value":"VSGBFURZPB65","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_16","alias_value":"VSGBFURZPB65MFCR","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_8","alias_value":"VSGBFURZ","created_at":"2026-07-05T00:58:29Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2020:VSGBFURZPB65MFCRCIIRAN3EZA","target":"record","payload":{"canonical_record":{"source":{"id":"2004.12993","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-27T17:58:05Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"1bbd1d03b192a4255bb12439d73db6d9410c00fcd3e07597e287abf41e934e8e","abstract_canon_sha256":"23e10bbf062d188fdb9f6ad92788bb2d3ddcc1e69ccae94a6fba1ca60af2b324"},"schema_version":"1.0"},"canonical_sha256":"ac8c12d239787dd614511211103764c827c80c977ff83cf689833bfe586fb6c8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T00:58:29.305727Z","signature_b64":"x+dB6t3D+Qu7KH9PfXyacQcmmDxaOaVpCrQqjlC7TiGZHKEl28u34NOXxawUJnPDVXetLgWUqiVdYcKT9xmWCg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"ac8c12d239787dd614511211103764c827c80c977ff83cf689833bfe586fb6c8","last_reissued_at":"2026-07-05T00:58:29.305220Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T00:58:29.305220Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2004.12993","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":"2026-07-05T00:58:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IyL0PU51NNYkSwQ/Ml5V50gGF0NHSYX3eMq/+weWhSREEN0X7+t1JPhz+hQjIHks4QoS8G+a0p6zCap+cv7pBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T02:11:14.192738Z"},"content_sha256":"12319277fc2fb54f33f3f3d824c5a1ecd446b63a9f1a2c63e76a8306c7f407b7","schema_version":"1.0","event_id":"sha256:12319277fc2fb54f33f3f3d824c5a1ecd446b63a9f1a2c63e76a8306c7f407b7"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2020:VSGBFURZPB65MFCRCIIRAN3EZA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Jaejun Lee, Jimmy Lin, Ji Xin, Raphael Tang, Yaoliang Yu","submitted_at":"2020-04-27T17:58:05Z","abstract_excerpt":"Large-scale pre-trained language models such as BERT have brought significant improvements to NLP applications. However, they are also notorious for being slow in inference, which makes them difficult to deploy in real-time applications. We propose a simple but effective method, DeeBERT, to accelerate BERT inference. Our approach allows samples to exit earlier without passing through the entire model. Experiments show that DeeBERT is able to save up to ~40% inference time with minimal degradation in model quality. Further analyses show different behaviors in the BERT transformer layers and als"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.12993","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/2004.12993/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":"2026-07-05T00:58:29Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"N2iusxg5j2vv5XErVy2V0HJjm83ILz/+CcyMNQefsetcZQ9ZpJbdR6M31H2Ks1Y1JjXoj+0W8xz2sQ9l1206Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T02:11:14.193722Z"},"content_sha256":"e06b4b5b5df434a40cb4fd5895cbe24cfac428e174fa2eb171bb39a46983d9a3","schema_version":"1.0","event_id":"sha256:e06b4b5b5df434a40cb4fd5895cbe24cfac428e174fa2eb171bb39a46983d9a3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/VSGBFURZPB65MFCRCIIRAN3EZA/bundle.json","state_url":"https://pith.science/pith/VSGBFURZPB65MFCRCIIRAN3EZA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/VSGBFURZPB65MFCRCIIRAN3EZA/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":"2026-08-06T02:11:14Z","links":{"resolver":"https://pith.science/pith/VSGBFURZPB65MFCRCIIRAN3EZA","bundle":"https://pith.science/pith/VSGBFURZPB65MFCRCIIRAN3EZA/bundle.json","state":"https://pith.science/pith/VSGBFURZPB65MFCRCIIRAN3EZA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/VSGBFURZPB65MFCRCIIRAN3EZA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2020:VSGBFURZPB65MFCRCIIRAN3EZA","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":"23e10bbf062d188fdb9f6ad92788bb2d3ddcc1e69ccae94a6fba1ca60af2b324","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-27T17:58:05Z","title_canon_sha256":"1bbd1d03b192a4255bb12439d73db6d9410c00fcd3e07597e287abf41e934e8e"},"schema_version":"1.0","source":{"id":"2004.12993","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2004.12993","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"arxiv_version","alias_value":"2004.12993v1","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2004.12993","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_12","alias_value":"VSGBFURZPB65","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_16","alias_value":"VSGBFURZPB65MFCR","created_at":"2026-07-05T00:58:29Z"},{"alias_kind":"pith_short_8","alias_value":"VSGBFURZ","created_at":"2026-07-05T00:58:29Z"}],"graph_snapshots":[{"event_id":"sha256:e06b4b5b5df434a40cb4fd5895cbe24cfac428e174fa2eb171bb39a46983d9a3","target":"graph","created_at":"2026-07-05T00:58:29Z","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/2004.12993/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Large-scale pre-trained language models such as BERT have brought significant improvements to NLP applications. However, they are also notorious for being slow in inference, which makes them difficult to deploy in real-time applications. We propose a simple but effective method, DeeBERT, to accelerate BERT inference. Our approach allows samples to exit earlier without passing through the entire model. Experiments show that DeeBERT is able to save up to ~40% inference time with minimal degradation in model quality. Further analyses show different behaviors in the BERT transformer layers and als","authors_text":"Jaejun Lee, Jimmy Lin, Ji Xin, Raphael Tang, Yaoliang Yu","cross_cats":["cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-27T17:58:05Z","title":"DeeBERT: Dynamic Early Exiting for Accelerating BERT Inference"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2004.12993","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":"sha256:12319277fc2fb54f33f3f3d824c5a1ecd446b63a9f1a2c63e76a8306c7f407b7","target":"record","created_at":"2026-07-05T00:58:29Z","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":"23e10bbf062d188fdb9f6ad92788bb2d3ddcc1e69ccae94a6fba1ca60af2b324","cross_cats_sorted":["cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2020-04-27T17:58:05Z","title_canon_sha256":"1bbd1d03b192a4255bb12439d73db6d9410c00fcd3e07597e287abf41e934e8e"},"schema_version":"1.0","source":{"id":"2004.12993","kind":"arxiv","version":1}},"canonical_sha256":"ac8c12d239787dd614511211103764c827c80c977ff83cf689833bfe586fb6c8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"ac8c12d239787dd614511211103764c827c80c977ff83cf689833bfe586fb6c8","first_computed_at":"2026-07-05T00:58:29.305220Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T00:58:29.305220Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"x+dB6t3D+Qu7KH9PfXyacQcmmDxaOaVpCrQqjlC7TiGZHKEl28u34NOXxawUJnPDVXetLgWUqiVdYcKT9xmWCg==","signature_status":"signed_v1","signed_at":"2026-07-05T00:58:29.305727Z","signed_message":"canonical_sha256_bytes"},"source_id":"2004.12993","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:12319277fc2fb54f33f3f3d824c5a1ecd446b63a9f1a2c63e76a8306c7f407b7","sha256:e06b4b5b5df434a40cb4fd5895cbe24cfac428e174fa2eb171bb39a46983d9a3"],"state_sha256":"d35f5f8604eda2c6c0cb882dc3cdf4127ea803c8da77f3f7360f7f717d53c45b"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KaQFayC6TEJjJsUFL2NfapEVyeA3fk/B+U0trQQuhXuJxPG1wHlK1gerFpaKs/rpH3+qJZtY9SB/+N1JzHsrAw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T02:11:14.199631Z","bundle_sha256":"e1b63d8ea1b1cd438165f3f4b1c519ea531f61120c4a1695fe4d0f94e698bc05"}}