{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:B3H2E7XHYT7QLP7GJLACML6N6G","short_pith_number":"pith:B3H2E7XH","canonical_record":{"source":{"id":"2505.07610","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-12T14:31:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"546c4dee6c424041bb9c120c519da3e89019efe40b50718b695384b8f299928a","abstract_canon_sha256":"d2998f8c8e77fb5eac06f7252ec4dc859983b2f94fd6fd7985c9d79cfd570464"},"schema_version":"1.0"},"canonical_sha256":"0ecfa27ee7c4ff05bfe64ac0262fcdf1b1c5a751797c6c17bfa5585b11db9764","source":{"kind":"arxiv","id":"2505.07610","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.07610","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"arxiv_version","alias_value":"2505.07610v2","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.07610","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_12","alias_value":"B3H2E7XHYT7Q","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_16","alias_value":"B3H2E7XHYT7QLP7G","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_8","alias_value":"B3H2E7XH","created_at":"2026-07-05T11:05:24Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:B3H2E7XHYT7QLP7GJLACML6N6G","target":"record","payload":{"canonical_record":{"source":{"id":"2505.07610","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-12T14:31:51Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"546c4dee6c424041bb9c120c519da3e89019efe40b50718b695384b8f299928a","abstract_canon_sha256":"d2998f8c8e77fb5eac06f7252ec4dc859983b2f94fd6fd7985c9d79cfd570464"},"schema_version":"1.0"},"canonical_sha256":"0ecfa27ee7c4ff05bfe64ac0262fcdf1b1c5a751797c6c17bfa5585b11db9764","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:05:24.804851Z","signature_b64":"lj7QqtNsMtrPBRf2/cWkpa9rygFuf0FAE3xgZ20WTkywGd0r8wMcHbtw0VFMWa3/zJxe0UyjCQW6KpTDfDwjBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"0ecfa27ee7c4ff05bfe64ac0262fcdf1b1c5a751797c6c17bfa5585b11db9764","last_reissued_at":"2026-07-05T11:05:24.804381Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:05:24.804381Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2505.07610","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:05:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MZG4p5vLPpIU5CE+eMn+FXKog9wq1T/aFt8ruXiRdqwU2MsSItXw6uN9SR1CXFhoKeKcFhXF9Qw5rypKQOd+Cw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T18:42:01.400099Z"},"content_sha256":"34275b354e8a037c09ed9e564e2851b8ce3f767851e13328f192c215780f26af","schema_version":"1.0","event_id":"sha256:34275b354e8a037c09ed9e564e2851b8ce3f767851e13328f192c215780f26af"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:B3H2E7XHYT7QLP7GJLACML6N6G","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Concept-Level Explainability for Auditing & Steering LLM Responses","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Kenza Amara, Mennatallah El-Assady, Rita Sevastjanova","submitted_at":"2025-05-12T14:31:51Z","abstract_excerpt":"As large language models (LLMs) become widely deployed, concerns about their safety and alignment grow. An approach to steer LLM behavior, such as mitigating biases or defending against jailbreaks, is to identify which parts of a prompt influence specific aspects of the model's output. Token-level attribution methods offer a promising solution, but still struggle in text generation, explaining the presence of each token in the output separately, rather than the underlying semantics of the entire LLM response. We introduce ConceptX, a model-agnostic, concept-level explainability method that ide"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.07610","kind":"arxiv","version":2},"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/2505.07610/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":"2026-07-05T11:05:24Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"KEutQ7fewDsL/wUYY0zerrBMlPhjGw1VtNar8xg2PhnpbJMP4FG7rNY4/ln2P/L9y0mGzpTsityTe7Xmg0XiDw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T18:42:01.401009Z"},"content_sha256":"b1bd6a4648af227bd35fda5f515ce8b6ca290ce1aad695305609f141c4e23b5f","schema_version":"1.0","event_id":"sha256:b1bd6a4648af227bd35fda5f515ce8b6ca290ce1aad695305609f141c4e23b5f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/B3H2E7XHYT7QLP7GJLACML6N6G/bundle.json","state_url":"https://pith.science/pith/B3H2E7XHYT7QLP7GJLACML6N6G/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/B3H2E7XHYT7QLP7GJLACML6N6G/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":"2026-08-14T18:42:01Z","links":{"resolver":"https://pith.science/pith/B3H2E7XHYT7QLP7GJLACML6N6G","bundle":"https://pith.science/pith/B3H2E7XHYT7QLP7GJLACML6N6G/bundle.json","state":"https://pith.science/pith/B3H2E7XHYT7QLP7GJLACML6N6G/state.json","well_known_bundle":"https://pith.science/.well-known/pith/B3H2E7XHYT7QLP7GJLACML6N6G/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:B3H2E7XHYT7QLP7GJLACML6N6G","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":"d2998f8c8e77fb5eac06f7252ec4dc859983b2f94fd6fd7985c9d79cfd570464","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-12T14:31:51Z","title_canon_sha256":"546c4dee6c424041bb9c120c519da3e89019efe40b50718b695384b8f299928a"},"schema_version":"1.0","source":{"id":"2505.07610","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2505.07610","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"arxiv_version","alias_value":"2505.07610v2","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2505.07610","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_12","alias_value":"B3H2E7XHYT7Q","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_16","alias_value":"B3H2E7XHYT7QLP7G","created_at":"2026-07-05T11:05:24Z"},{"alias_kind":"pith_short_8","alias_value":"B3H2E7XH","created_at":"2026-07-05T11:05:24Z"}],"graph_snapshots":[{"event_id":"sha256:b1bd6a4648af227bd35fda5f515ce8b6ca290ce1aad695305609f141c4e23b5f","target":"graph","created_at":"2026-07-05T11:05:24Z","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/2505.07610/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"As large language models (LLMs) become widely deployed, concerns about their safety and alignment grow. An approach to steer LLM behavior, such as mitigating biases or defending against jailbreaks, is to identify which parts of a prompt influence specific aspects of the model's output. Token-level attribution methods offer a promising solution, but still struggle in text generation, explaining the presence of each token in the output separately, rather than the underlying semantics of the entire LLM response. We introduce ConceptX, a model-agnostic, concept-level explainability method that ide","authors_text":"Kenza Amara, Mennatallah El-Assady, Rita Sevastjanova","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-12T14:31:51Z","title":"Concept-Level Explainability for Auditing & Steering LLM Responses"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2505.07610","kind":"arxiv","version":2},"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":"sha256:34275b354e8a037c09ed9e564e2851b8ce3f767851e13328f192c215780f26af","target":"record","created_at":"2026-07-05T11:05:24Z","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":"d2998f8c8e77fb5eac06f7252ec4dc859983b2f94fd6fd7985c9d79cfd570464","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2025-05-12T14:31:51Z","title_canon_sha256":"546c4dee6c424041bb9c120c519da3e89019efe40b50718b695384b8f299928a"},"schema_version":"1.0","source":{"id":"2505.07610","kind":"arxiv","version":2}},"canonical_sha256":"0ecfa27ee7c4ff05bfe64ac0262fcdf1b1c5a751797c6c17bfa5585b11db9764","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"0ecfa27ee7c4ff05bfe64ac0262fcdf1b1c5a751797c6c17bfa5585b11db9764","first_computed_at":"2026-07-05T11:05:24.804381Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:05:24.804381Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lj7QqtNsMtrPBRf2/cWkpa9rygFuf0FAE3xgZ20WTkywGd0r8wMcHbtw0VFMWa3/zJxe0UyjCQW6KpTDfDwjBg==","signature_status":"signed_v1","signed_at":"2026-07-05T11:05:24.804851Z","signed_message":"canonical_sha256_bytes"},"source_id":"2505.07610","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:34275b354e8a037c09ed9e564e2851b8ce3f767851e13328f192c215780f26af","sha256:b1bd6a4648af227bd35fda5f515ce8b6ca290ce1aad695305609f141c4e23b5f"],"state_sha256":"43ec359dc24ab2c8b0ebb095f199b8fdb0d8acd8d74d3639f6f6e33e45ca00a3"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Vrgclv8foA/g6DFT0532qyjM1s+0sBQimvEVBl0NB9oxVV/xAe9nS+Kv3QU2BBlj9mnUNI5vLYwVlR/VZXx4CQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T18:42:01.408091Z","bundle_sha256":"0508903ae0765e19b12f1d2d06d7468d51839d9bb7adf04cb7c16c2a372e5764"}}