{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:LNRTXYNCJ35YOAJDBEWQKIY2YB","short_pith_number":"pith:LNRTXYNC","canonical_record":{"source":{"id":"2410.21159","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-10-28T15:59:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0486fee37f3f09e10baa1db24388af1b805b926b01a2a131192078d63f450a4d","abstract_canon_sha256":"ac7d3b3a0ede06a9a101a1506afa34e59b9960d2f5fce57b4cba4f66acf0ffe1"},"schema_version":"1.0"},"canonical_sha256":"5b633be1a24efb870123092d05231ac05b4668699ce065345f15a1c691ae3bf8","source":{"kind":"arxiv","id":"2410.21159","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.21159","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"arxiv_version","alias_value":"2410.21159v2","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21159","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_12","alias_value":"LNRTXYNCJ35Y","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_16","alias_value":"LNRTXYNCJ35YOAJD","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_8","alias_value":"LNRTXYNC","created_at":"2026-07-05T10:07:02Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:LNRTXYNCJ35YOAJDBEWQKIY2YB","target":"record","payload":{"canonical_record":{"source":{"id":"2410.21159","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-10-28T15:59:31Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"0486fee37f3f09e10baa1db24388af1b805b926b01a2a131192078d63f450a4d","abstract_canon_sha256":"ac7d3b3a0ede06a9a101a1506afa34e59b9960d2f5fce57b4cba4f66acf0ffe1"},"schema_version":"1.0"},"canonical_sha256":"5b633be1a24efb870123092d05231ac05b4668699ce065345f15a1c691ae3bf8","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:07:02.124348Z","signature_b64":"jtn2mbyiC9UHb/OpxlA/4KOrL+jhWvgTqf5CXTVeuVa72yxyciN21bsqJKML7oo235hw18B1Z4xjvNn3D+AmAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"5b633be1a24efb870123092d05231ac05b4668699ce065345f15a1c691ae3bf8","last_reissued_at":"2026-07-05T10:07:02.123926Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:07:02.123926Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2410.21159","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":"2026-07-05T10:07:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"aizFznUOKjiPUjx6XEHSIRW13qgIot7i4qoVB6lh/uYL3ZUv+Y2HCSd8kqavUVWgomIPi9ACrK+UzcymZeE7Dw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T18:29:23.642941Z"},"content_sha256":"269a26e1e2a50f05281531001c64aff9912828f74e1b9e828b761279da58857d","schema_version":"1.0","event_id":"sha256:269a26e1e2a50f05281531001c64aff9912828f74e1b9e828b761279da58857d"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:LNRTXYNCJ35YOAJDBEWQKIY2YB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"CURATe: Benchmarking Personalised Alignment of Conversational AI Assistants","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.HC","authors_text":"Andrei Lupu, Benjamin Ellis, Jakob Foerster, Lize Alberts","submitted_at":"2024-10-28T15:59:31Z","abstract_excerpt":"We introduce a multi-turn benchmark for evaluating personalised alignment in LLM-based AI assistants, focusing on their ability to handle user-provided safety-critical contexts. Our assessment of ten leading models across five scenarios (with 337 use cases each) reveals systematic inconsistencies in maintaining user-specific consideration, with even top-rated \"harmless\" models making recommendations that should be recognised as obviously harmful to the user given the context provided. Key failure modes include inappropriate weighing of conflicting preferences, sycophancy (prioritising desires "},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21159","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/2410.21159/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":"2026-07-05T10:07:02Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"akXVJWHIch1u9gu7rt7ZZI6FdY7qxfkQ1iOXg2d5F5hx0U+oHXMUSEzYQQOc85MwDXHWmgoIz2uAbJQ+16PFAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-09T18:29:23.645889Z"},"content_sha256":"a1e6f18273ccee218ed8bca22a1027736884584857816328bdcdfad8842718cb","schema_version":"1.0","event_id":"sha256:a1e6f18273ccee218ed8bca22a1027736884584857816328bdcdfad8842718cb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/bundle.json","state_url":"https://pith.science/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/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":"2026-08-09T18:29:23Z","links":{"resolver":"https://pith.science/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB","bundle":"https://pith.science/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/bundle.json","state":"https://pith.science/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/LNRTXYNCJ35YOAJDBEWQKIY2YB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:LNRTXYNCJ35YOAJDBEWQKIY2YB","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":"ac7d3b3a0ede06a9a101a1506afa34e59b9960d2f5fce57b4cba4f66acf0ffe1","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-10-28T15:59:31Z","title_canon_sha256":"0486fee37f3f09e10baa1db24388af1b805b926b01a2a131192078d63f450a4d"},"schema_version":"1.0","source":{"id":"2410.21159","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2410.21159","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"arxiv_version","alias_value":"2410.21159v2","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.21159","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_12","alias_value":"LNRTXYNCJ35Y","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_16","alias_value":"LNRTXYNCJ35YOAJD","created_at":"2026-07-05T10:07:02Z"},{"alias_kind":"pith_short_8","alias_value":"LNRTXYNC","created_at":"2026-07-05T10:07:02Z"}],"graph_snapshots":[{"event_id":"sha256:a1e6f18273ccee218ed8bca22a1027736884584857816328bdcdfad8842718cb","target":"graph","created_at":"2026-07-05T10:07:02Z","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/2410.21159/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"We introduce a multi-turn benchmark for evaluating personalised alignment in LLM-based AI assistants, focusing on their ability to handle user-provided safety-critical contexts. Our assessment of ten leading models across five scenarios (with 337 use cases each) reveals systematic inconsistencies in maintaining user-specific consideration, with even top-rated \"harmless\" models making recommendations that should be recognised as obviously harmful to the user given the context provided. Key failure modes include inappropriate weighing of conflicting preferences, sycophancy (prioritising desires ","authors_text":"Andrei Lupu, Benjamin Ellis, Jakob Foerster, Lize Alberts","cross_cats":["cs.AI"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-10-28T15:59:31Z","title":"CURATe: Benchmarking Personalised Alignment of Conversational AI Assistants"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.21159","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":"sha256:269a26e1e2a50f05281531001c64aff9912828f74e1b9e828b761279da58857d","target":"record","created_at":"2026-07-05T10:07:02Z","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":"ac7d3b3a0ede06a9a101a1506afa34e59b9960d2f5fce57b4cba4f66acf0ffe1","cross_cats_sorted":["cs.AI"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.HC","submitted_at":"2024-10-28T15:59:31Z","title_canon_sha256":"0486fee37f3f09e10baa1db24388af1b805b926b01a2a131192078d63f450a4d"},"schema_version":"1.0","source":{"id":"2410.21159","kind":"arxiv","version":2}},"canonical_sha256":"5b633be1a24efb870123092d05231ac05b4668699ce065345f15a1c691ae3bf8","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"5b633be1a24efb870123092d05231ac05b4668699ce065345f15a1c691ae3bf8","first_computed_at":"2026-07-05T10:07:02.123926Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:07:02.123926Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"jtn2mbyiC9UHb/OpxlA/4KOrL+jhWvgTqf5CXTVeuVa72yxyciN21bsqJKML7oo235hw18B1Z4xjvNn3D+AmAA==","signature_status":"signed_v1","signed_at":"2026-07-05T10:07:02.124348Z","signed_message":"canonical_sha256_bytes"},"source_id":"2410.21159","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:269a26e1e2a50f05281531001c64aff9912828f74e1b9e828b761279da58857d","sha256:a1e6f18273ccee218ed8bca22a1027736884584857816328bdcdfad8842718cb"],"state_sha256":"5cb1b3e2dcda3f47cf259b1820cdf55ead381672d9b53e11f6641228905f22b9"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3wnFC2yYj1BfEV/TP6Spdz3cMluDEAL25A9yl39Ev3MfvVsjM1iTpWB5Ck6gT5+HjbpNbPKHhzKGyp5i6rqADA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-09T18:29:23.653171Z","bundle_sha256":"f7765f0e9d54f45853bc8012ca666c4df8aa8019aaf1c9c88b4d1eca2f0d593f"}}