{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:SSAUXM4FORN4FCEFM2VMUM4XD3","short_pith_number":"pith:SSAUXM4F","canonical_record":{"source":{"id":"2404.08690","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T22:56:05Z","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"title_canon_sha256":"83f7932dcf5a13ac6f6144df6c3ad5457c0138fdf06d472099c1d9e1a659765f","abstract_canon_sha256":"84254b8aa0f62fcf017c0a80df5212fe5ba814110d6d6b9308b9c729f676824e"},"schema_version":"1.0"},"canonical_sha256":"94814bb385745bc2888566aaca33971ee4ffcb9a05b59415c659fdef66227b9a","source":{"kind":"arxiv","id":"2404.08690","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.08690","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"arxiv_version","alias_value":"2404.08690v1","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.08690","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_12","alias_value":"SSAUXM4FORN4","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_16","alias_value":"SSAUXM4FORN4FCEF","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_8","alias_value":"SSAUXM4F","created_at":"2026-07-05T08:07:38Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:SSAUXM4FORN4FCEFM2VMUM4XD3","target":"record","payload":{"canonical_record":{"source":{"id":"2404.08690","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T22:56:05Z","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"title_canon_sha256":"83f7932dcf5a13ac6f6144df6c3ad5457c0138fdf06d472099c1d9e1a659765f","abstract_canon_sha256":"84254b8aa0f62fcf017c0a80df5212fe5ba814110d6d6b9308b9c729f676824e"},"schema_version":"1.0"},"canonical_sha256":"94814bb385745bc2888566aaca33971ee4ffcb9a05b59415c659fdef66227b9a","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:07:38.250075Z","signature_b64":"rWRUq/L5CA9Nd5Ipz9cHjZOm53qI4vUsSNWZ3SkRVOvLqU+UzDyS5WlCPfFyj3rwkONVbUFkfidDtTnu8+LICA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"94814bb385745bc2888566aaca33971ee4ffcb9a05b59415c659fdef66227b9a","last_reissued_at":"2026-07-05T08:07:38.249634Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:07:38.249634Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2404.08690","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":"2026-07-05T08:07:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"FnR9MXa5tZq6V4w3i8gsu2jylFs9+dOiXFNSVd8MZSvSkSYKdyvQgXOdhiD6MiynwP8yqpZ9KCXbqNIcH31NAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:33:58.223849Z"},"content_sha256":"a9bda6be47f02ce88edc33355c8ce9574e456b8f9b3e880d2d181d8e6c250af0","schema_version":"1.0","event_id":"sha256:a9bda6be47f02ce88edc33355c8ce9574e456b8f9b3e880d2d181d8e6c250af0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:SSAUXM4FORN4FCEFM2VMUM4XD3","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Towards Building a Robust Toxicity Predictor","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI","cs.CR","cs.LG"],"primary_cat":"cs.CL","authors_text":"Dmitriy Bespalov, Liutong Zhou, Sourav Bhabesh, Yanjun Qi, Yi Xiang","submitted_at":"2024-04-09T22:56:05Z","abstract_excerpt":"Recent NLP literature pays little attention to the robustness of toxicity language predictors, while these systems are most likely to be used in adversarial contexts. This paper presents a novel adversarial attack, \\texttt{ToxicTrap}, introducing small word-level perturbations to fool SOTA text classifiers to predict toxic text samples as benign. ToxicTrap exploits greedy based search strategies to enable fast and effective generation of toxic adversarial examples. Two novel goal function designs allow ToxicTrap to identify weaknesses in both multiclass and multilabel toxic language detectors."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.08690","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/2404.08690/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":"2026-07-05T08:07:38Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MSCd/iTaaqrAzXy3QXTSbNmaHEC0w8O3KpozDPtCN86i6z8ZTJReGZlcSek2ECifNGcQp39yJOtAAeZrzJ2DBw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T00:33:58.224396Z"},"content_sha256":"1e1c75508b30789cc48e972e852347748a208f3c62e422767f123046c2f5cdd7","schema_version":"1.0","event_id":"sha256:1e1c75508b30789cc48e972e852347748a208f3c62e422767f123046c2f5cdd7"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/bundle.json","state_url":"https://pith.science/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/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":"2026-08-09T00:33:58Z","links":{"resolver":"https://pith.science/pith/SSAUXM4FORN4FCEFM2VMUM4XD3","bundle":"https://pith.science/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/bundle.json","state":"https://pith.science/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/state.json","well_known_bundle":"https://pith.science/.well-known/pith/SSAUXM4FORN4FCEFM2VMUM4XD3/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:SSAUXM4FORN4FCEFM2VMUM4XD3","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":"84254b8aa0f62fcf017c0a80df5212fe5ba814110d6d6b9308b9c729f676824e","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T22:56:05Z","title_canon_sha256":"83f7932dcf5a13ac6f6144df6c3ad5457c0138fdf06d472099c1d9e1a659765f"},"schema_version":"1.0","source":{"id":"2404.08690","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2404.08690","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"arxiv_version","alias_value":"2404.08690v1","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2404.08690","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_12","alias_value":"SSAUXM4FORN4","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_16","alias_value":"SSAUXM4FORN4FCEF","created_at":"2026-07-05T08:07:38Z"},{"alias_kind":"pith_short_8","alias_value":"SSAUXM4F","created_at":"2026-07-05T08:07:38Z"}],"graph_snapshots":[{"event_id":"sha256:1e1c75508b30789cc48e972e852347748a208f3c62e422767f123046c2f5cdd7","target":"graph","created_at":"2026-07-05T08:07:38Z","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/2404.08690/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Recent NLP literature pays little attention to the robustness of toxicity language predictors, while these systems are most likely to be used in adversarial contexts. This paper presents a novel adversarial attack, \\texttt{ToxicTrap}, introducing small word-level perturbations to fool SOTA text classifiers to predict toxic text samples as benign. ToxicTrap exploits greedy based search strategies to enable fast and effective generation of toxic adversarial examples. Two novel goal function designs allow ToxicTrap to identify weaknesses in both multiclass and multilabel toxic language detectors.","authors_text":"Dmitriy Bespalov, Liutong Zhou, Sourav Bhabesh, Yanjun Qi, Yi Xiang","cross_cats":["cs.AI","cs.CR","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T22:56:05Z","title":"Towards Building a Robust Toxicity Predictor"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2404.08690","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:a9bda6be47f02ce88edc33355c8ce9574e456b8f9b3e880d2d181d8e6c250af0","target":"record","created_at":"2026-07-05T08:07:38Z","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":"84254b8aa0f62fcf017c0a80df5212fe5ba814110d6d6b9308b9c729f676824e","cross_cats_sorted":["cs.AI","cs.CR","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-04-09T22:56:05Z","title_canon_sha256":"83f7932dcf5a13ac6f6144df6c3ad5457c0138fdf06d472099c1d9e1a659765f"},"schema_version":"1.0","source":{"id":"2404.08690","kind":"arxiv","version":1}},"canonical_sha256":"94814bb385745bc2888566aaca33971ee4ffcb9a05b59415c659fdef66227b9a","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"94814bb385745bc2888566aaca33971ee4ffcb9a05b59415c659fdef66227b9a","first_computed_at":"2026-07-05T08:07:38.249634Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:07:38.249634Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"rWRUq/L5CA9Nd5Ipz9cHjZOm53qI4vUsSNWZ3SkRVOvLqU+UzDyS5WlCPfFyj3rwkONVbUFkfidDtTnu8+LICA==","signature_status":"signed_v1","signed_at":"2026-07-05T08:07:38.250075Z","signed_message":"canonical_sha256_bytes"},"source_id":"2404.08690","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:a9bda6be47f02ce88edc33355c8ce9574e456b8f9b3e880d2d181d8e6c250af0","sha256:1e1c75508b30789cc48e972e852347748a208f3c62e422767f123046c2f5cdd7"],"state_sha256":"e39bd4e4f67dee384b45bc64ebd01176a8bb2d9ddc8a4909e3c4713e2d6d3c3a"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"i5DinD7LejG+UePXYKJ2G85eLCiI7FhkKSok98rJIfSwF/mAI22P91zmU6UttVMZhSclgw6oTTfCJXP/yXp7BA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T00:33:58.229414Z","bundle_sha256":"ed6681e87e862131375134c3587b6b0052281cc3b3a064bb5a4d3879f94f538a"}}