{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2019:PPN7QATKDJC3WGVENA5L3QV6HY","short_pith_number":"pith:PPN7QATK","canonical_record":{"source":{"id":"1905.10729","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T04:57:55Z","cross_cats_sorted":["cs.CR","cs.CV","stat.ML"],"title_canon_sha256":"591a5c3d8ded4a245d74165c6851d7c72cf341214449b455bf7f8ba7c3fb1cc9","abstract_canon_sha256":"71084db8d1d4e823328f9998424da7518aa576658c9b9c721d9c99fb5c8ce707"},"schema_version":"1.0"},"canonical_sha256":"7bdbf8026a1a45bb1aa4683abdc2be3e14ea17c46800b629a87a8c2cb9109178","source":{"kind":"arxiv","id":"1905.10729","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1905.10729","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"arxiv_version","alias_value":"1905.10729v1","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.10729","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"pith_short_12","alias_value":"PPN7QATKDJC3","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_16","alias_value":"PPN7QATKDJC3WGVE","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_8","alias_value":"PPN7QATK","created_at":"2026-05-18T12:33:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2019:PPN7QATKDJC3WGVENA5L3QV6HY","target":"record","payload":{"canonical_record":{"source":{"id":"1905.10729","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T04:57:55Z","cross_cats_sorted":["cs.CR","cs.CV","stat.ML"],"title_canon_sha256":"591a5c3d8ded4a245d74165c6851d7c72cf341214449b455bf7f8ba7c3fb1cc9","abstract_canon_sha256":"71084db8d1d4e823328f9998424da7518aa576658c9b9c721d9c99fb5c8ce707"},"schema_version":"1.0"},"canonical_sha256":"7bdbf8026a1a45bb1aa4683abdc2be3e14ea17c46800b629a87a8c2cb9109178","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:45:05.411358Z","signature_b64":"5mNIzWDojo0APRkX+HytbhIdDuupW8x+m4c6sYZvoxAWSeGAyj5FuxCj9l54DYKNIzubwCZdg+VI8Vwz0zjMAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7bdbf8026a1a45bb1aa4683abdc2be3e14ea17c46800b629a87a8c2cb9109178","last_reissued_at":"2026-05-17T23:45:05.410813Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:45:05.410813Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1905.10729","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":"2026-05-17T23:45:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"acEGMb8Pj+zpEOVo0Q7b/ACQEImdPS0BlcSmrny5sF1sbApMVGSRIqIK0Q1jeYcSN1oPqtMCGntaRVAzPBT3AQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T22:25:47.634034Z"},"content_sha256":"355f3723ba5a2f9936e016a02c1977e74dba43a918b18858ec2ffaae372280d3","schema_version":"1.0","event_id":"sha256:355f3723ba5a2f9936e016a02c1977e74dba43a918b18858ec2ffaae372280d3"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2019:PPN7QATKDJC3WGVENA5L3QV6HY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Purifying Adversarial Perturbation with Adversarially Trained Auto-encoders","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.CR","cs.CV","stat.ML"],"primary_cat":"cs.LG","authors_text":"Hebi Li, Jin Tian, Qi Xiao, Shixin Tian","submitted_at":"2019-05-26T04:57:55Z","abstract_excerpt":"Machine learning models are vulnerable to adversarial examples. Iterative adversarial training has shown promising results against strong white-box attacks. However, adversarial training is very expensive, and every time a model needs to be protected, such expensive training scheme needs to be performed. In this paper, we propose to apply iterative adversarial training scheme to an external auto-encoder, which once trained can be used to protect other models directly. We empirically show that our model outperforms other purifying-based methods against white-box attacks, and transfers well to d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1905.10729","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":""},"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":"2026-05-17T23:45:05Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"w4RK0xAeXbEbJGMmd8uTHp0GEM2DCVn7HTuhm1nvJrc2PWfuARfQz40zORpVJGXZzn6BDQSBQbxygTeOj62dAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-06T22:25:47.634743Z"},"content_sha256":"aafff00e0fdffb6a70186a24e0a4639a7a9bf6467acfd2ba94449db2a2e96130","schema_version":"1.0","event_id":"sha256:aafff00e0fdffb6a70186a24e0a4639a7a9bf6467acfd2ba94449db2a2e96130"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/PPN7QATKDJC3WGVENA5L3QV6HY/bundle.json","state_url":"https://pith.science/pith/PPN7QATKDJC3WGVENA5L3QV6HY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/PPN7QATKDJC3WGVENA5L3QV6HY/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":"2026-06-06T22:25:47Z","links":{"resolver":"https://pith.science/pith/PPN7QATKDJC3WGVENA5L3QV6HY","bundle":"https://pith.science/pith/PPN7QATKDJC3WGVENA5L3QV6HY/bundle.json","state":"https://pith.science/pith/PPN7QATKDJC3WGVENA5L3QV6HY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/PPN7QATKDJC3WGVENA5L3QV6HY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2019:PPN7QATKDJC3WGVENA5L3QV6HY","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":"71084db8d1d4e823328f9998424da7518aa576658c9b9c721d9c99fb5c8ce707","cross_cats_sorted":["cs.CR","cs.CV","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T04:57:55Z","title_canon_sha256":"591a5c3d8ded4a245d74165c6851d7c72cf341214449b455bf7f8ba7c3fb1cc9"},"schema_version":"1.0","source":{"id":"1905.10729","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1905.10729","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"arxiv_version","alias_value":"1905.10729v1","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1905.10729","created_at":"2026-05-17T23:45:05Z"},{"alias_kind":"pith_short_12","alias_value":"PPN7QATKDJC3","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_16","alias_value":"PPN7QATKDJC3WGVE","created_at":"2026-05-18T12:33:24Z"},{"alias_kind":"pith_short_8","alias_value":"PPN7QATK","created_at":"2026-05-18T12:33:24Z"}],"graph_snapshots":[{"event_id":"sha256:aafff00e0fdffb6a70186a24e0a4639a7a9bf6467acfd2ba94449db2a2e96130","target":"graph","created_at":"2026-05-17T23:45:05Z","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"},"paper":{"abstract_excerpt":"Machine learning models are vulnerable to adversarial examples. Iterative adversarial training has shown promising results against strong white-box attacks. However, adversarial training is very expensive, and every time a model needs to be protected, such expensive training scheme needs to be performed. In this paper, we propose to apply iterative adversarial training scheme to an external auto-encoder, which once trained can be used to protect other models directly. We empirically show that our model outperforms other purifying-based methods against white-box attacks, and transfers well to d","authors_text":"Hebi Li, Jin Tian, Qi Xiao, Shixin Tian","cross_cats":["cs.CR","cs.CV","stat.ML"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T04:57:55Z","title":"Purifying Adversarial Perturbation with Adversarially Trained Auto-encoders"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1905.10729","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":"sha256:355f3723ba5a2f9936e016a02c1977e74dba43a918b18858ec2ffaae372280d3","target":"record","created_at":"2026-05-17T23:45:05Z","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":"71084db8d1d4e823328f9998424da7518aa576658c9b9c721d9c99fb5c8ce707","cross_cats_sorted":["cs.CR","cs.CV","stat.ML"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2019-05-26T04:57:55Z","title_canon_sha256":"591a5c3d8ded4a245d74165c6851d7c72cf341214449b455bf7f8ba7c3fb1cc9"},"schema_version":"1.0","source":{"id":"1905.10729","kind":"arxiv","version":1}},"canonical_sha256":"7bdbf8026a1a45bb1aa4683abdc2be3e14ea17c46800b629a87a8c2cb9109178","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"7bdbf8026a1a45bb1aa4683abdc2be3e14ea17c46800b629a87a8c2cb9109178","first_computed_at":"2026-05-17T23:45:05.410813Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-17T23:45:05.410813Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"5mNIzWDojo0APRkX+HytbhIdDuupW8x+m4c6sYZvoxAWSeGAyj5FuxCj9l54DYKNIzubwCZdg+VI8Vwz0zjMAw==","signature_status":"signed_v1","signed_at":"2026-05-17T23:45:05.411358Z","signed_message":"canonical_sha256_bytes"},"source_id":"1905.10729","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:355f3723ba5a2f9936e016a02c1977e74dba43a918b18858ec2ffaae372280d3","sha256:aafff00e0fdffb6a70186a24e0a4639a7a9bf6467acfd2ba94449db2a2e96130"],"state_sha256":"fa7d57be873b01d45c5003e8b610ac1fdd46d072910468d2d3762274a3396a35"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"rjSOysdgIZ0e6ClaRTozy1ZByHDIZrPYDqdTFf/5r+AGJSQB1rtonGKnyAw1VuEj8xmvkg+hsmpCoSQP2MDLBw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-06T22:25:47.638705Z","bundle_sha256":"1a8dd69fa21f5d5e5180be7314b974483ffa47689363b7df55057d06824e56c2"}}