{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:QTPM2VYZUKZKHGBAHQ4RWM5MKE","short_pith_number":"pith:QTPM2VYZ","canonical_record":{"source":{"id":"2401.16332","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T17:38:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"99b6c325a10599f4055b604925a6cb3e00e5e82132731ad1c39847aaf8e49fce","abstract_canon_sha256":"0f389946cf4bf37eafb3472df7b24c990b3ca1436e2f420d446821d472cc7138"},"schema_version":"1.0"},"canonical_sha256":"84decd5719a2b2a398203c391b33ac5116711fcc975facb91af9719856fdb294","source":{"kind":"arxiv","id":"2401.16332","version":5},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.16332","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"arxiv_version","alias_value":"2401.16332v5","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.16332","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_12","alias_value":"QTPM2VYZUKZK","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_16","alias_value":"QTPM2VYZUKZKHGBA","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_8","alias_value":"QTPM2VYZ","created_at":"2026-07-05T11:09:57Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:QTPM2VYZUKZKHGBAHQ4RWM5MKE","target":"record","payload":{"canonical_record":{"source":{"id":"2401.16332","kind":"arxiv","version":5},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T17:38:14Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"99b6c325a10599f4055b604925a6cb3e00e5e82132731ad1c39847aaf8e49fce","abstract_canon_sha256":"0f389946cf4bf37eafb3472df7b24c990b3ca1436e2f420d446821d472cc7138"},"schema_version":"1.0"},"canonical_sha256":"84decd5719a2b2a398203c391b33ac5116711fcc975facb91af9719856fdb294","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T11:09:57.661050Z","signature_b64":"LDmnnhyB438oJC0fgw8DqCrdsmHL/XUEMe317ZCfmSrpRBtd27wQfUnAtXmixC1ADizv3szlz7j6xlH1d8ZpBA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"84decd5719a2b2a398203c391b33ac5116711fcc975facb91af9719856fdb294","last_reissued_at":"2026-07-05T11:09:57.660554Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T11:09:57.660554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2401.16332","source_version":5,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T11:09:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"zpj9BNEBaX05EWCbrlXGrw9L8n3W3iqcV/S1DvIyhWMMhcFw3nOSesF6fkpzVGEtP9FDP1cprNMDBVNI36AkAA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T16:59:09.536670Z"},"content_sha256":"60bc03b0e8d97c560154c5ebdbdc9c73b445bd24ff5b57115e0089ad2e383d75","schema_version":"1.0","event_id":"sha256:60bc03b0e8d97c560154c5ebdbdc9c73b445bd24ff5b57115e0089ad2e383d75"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:QTPM2VYZUKZKHGBAHQ4RWM5MKE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Tradeoffs Between Alignment and Helpfulness in Language Models with Steering Methods","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amnon Shashua, Binyamin Rothberg, Dorin Shteyman, Noam Wies, Yoav Levine, Yotam Wolf","submitted_at":"2024-01-29T17:38:14Z","abstract_excerpt":"Language model alignment has become an important component of AI safety, allowing safe interactions between humans and language models, by enhancing desired behaviors and inhibiting undesired ones. It is often done by tuning the model or inserting preset aligning prompts. Recently, representation engineering, a method which alters the model's behavior via changing its representations post-training, was shown to be effective in aligning LLMs (Zou et al., 2023a). Representation engineering yields gains in alignment oriented tasks such as resistance to adversarial attacks and reduction of social "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.16332","kind":"arxiv","version":5},"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/2401.16332/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":"2026-07-05T11:09:57Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"LemaqZQ74DqIZKDRkghSva/xHlOKeiFqlUrxvQcrriQ/sPnu9rmW4I2TeD4e0EGz4Bi1xWQWkAGjFS+RRaHJAQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-12T16:59:09.537552Z"},"content_sha256":"38c863998a8fb2a72a38649ece8823c108a7c5a53888888a05bf9b5f0e1c751d","schema_version":"1.0","event_id":"sha256:38c863998a8fb2a72a38649ece8823c108a7c5a53888888a05bf9b5f0e1c751d"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/bundle.json","state_url":"https://pith.science/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/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":"2026-08-12T16:59:09Z","links":{"resolver":"https://pith.science/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE","bundle":"https://pith.science/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/bundle.json","state":"https://pith.science/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/QTPM2VYZUKZKHGBAHQ4RWM5MKE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:QTPM2VYZUKZKHGBAHQ4RWM5MKE","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":"0f389946cf4bf37eafb3472df7b24c990b3ca1436e2f420d446821d472cc7138","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T17:38:14Z","title_canon_sha256":"99b6c325a10599f4055b604925a6cb3e00e5e82132731ad1c39847aaf8e49fce"},"schema_version":"1.0","source":{"id":"2401.16332","kind":"arxiv","version":5}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2401.16332","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"arxiv_version","alias_value":"2401.16332v5","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2401.16332","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_12","alias_value":"QTPM2VYZUKZK","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_16","alias_value":"QTPM2VYZUKZKHGBA","created_at":"2026-07-05T11:09:57Z"},{"alias_kind":"pith_short_8","alias_value":"QTPM2VYZ","created_at":"2026-07-05T11:09:57Z"}],"graph_snapshots":[{"event_id":"sha256:38c863998a8fb2a72a38649ece8823c108a7c5a53888888a05bf9b5f0e1c751d","target":"graph","created_at":"2026-07-05T11:09:57Z","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/2401.16332/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Language model alignment has become an important component of AI safety, allowing safe interactions between humans and language models, by enhancing desired behaviors and inhibiting undesired ones. It is often done by tuning the model or inserting preset aligning prompts. Recently, representation engineering, a method which alters the model's behavior via changing its representations post-training, was shown to be effective in aligning LLMs (Zou et al., 2023a). Representation engineering yields gains in alignment oriented tasks such as resistance to adversarial attacks and reduction of social ","authors_text":"Amnon Shashua, Binyamin Rothberg, Dorin Shteyman, Noam Wies, Yoav Levine, Yotam Wolf","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T17:38:14Z","title":"Tradeoffs Between Alignment and Helpfulness in Language Models with Steering Methods"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2401.16332","kind":"arxiv","version":5},"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":"sha256:60bc03b0e8d97c560154c5ebdbdc9c73b445bd24ff5b57115e0089ad2e383d75","target":"record","created_at":"2026-07-05T11:09:57Z","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":"0f389946cf4bf37eafb3472df7b24c990b3ca1436e2f420d446821d472cc7138","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2024-01-29T17:38:14Z","title_canon_sha256":"99b6c325a10599f4055b604925a6cb3e00e5e82132731ad1c39847aaf8e49fce"},"schema_version":"1.0","source":{"id":"2401.16332","kind":"arxiv","version":5}},"canonical_sha256":"84decd5719a2b2a398203c391b33ac5116711fcc975facb91af9719856fdb294","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"84decd5719a2b2a398203c391b33ac5116711fcc975facb91af9719856fdb294","first_computed_at":"2026-07-05T11:09:57.660554Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T11:09:57.660554Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"LDmnnhyB438oJC0fgw8DqCrdsmHL/XUEMe317ZCfmSrpRBtd27wQfUnAtXmixC1ADizv3szlz7j6xlH1d8ZpBA==","signature_status":"signed_v1","signed_at":"2026-07-05T11:09:57.661050Z","signed_message":"canonical_sha256_bytes"},"source_id":"2401.16332","source_kind":"arxiv","source_version":5}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:60bc03b0e8d97c560154c5ebdbdc9c73b445bd24ff5b57115e0089ad2e383d75","sha256:38c863998a8fb2a72a38649ece8823c108a7c5a53888888a05bf9b5f0e1c751d"],"state_sha256":"04c5c19d2d31b2f0cfa6048c36d3be10bca2f499a5063375c70ebecf598759e6"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"uEMSYQ0lOAUH1BfYelPlf5OOnImrbj8jgchM2zdcbpEqz1wKbXHvswJX8QtR8PanCIuQ3oPC0dGFRUFxh1UaBA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-12T16:59:09.543104Z","bundle_sha256":"5a63db95ba7d30a808a7e941950cbfd6c08020fc7000e025aa77ee5de6011d9e"}}