{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2022:ZPXHUXBLB6PN77NKG5JK4KIIXB","short_pith_number":"pith:ZPXHUXBL","canonical_record":{"source":{"id":"2203.13553","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-25T10:19:35Z","cross_cats_sorted":[],"title_canon_sha256":"c9c6852b9ee495912ef8300870d3438748f1c336e2bc3d0caac0c67983974fff","abstract_canon_sha256":"7685008200a6d15d446e56e1f2f32842022cda30fd687b5be99e862e2e90df42"},"schema_version":"1.0"},"canonical_sha256":"cbee7a5c2b0f9edffdaa3752ae2908b852b2bf60038c06737493d8cd3c9998ca","source":{"kind":"arxiv","id":"2203.13553","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2203.13553","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"arxiv_version","alias_value":"2203.13553v1","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.13553","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_12","alias_value":"ZPXHUXBLB6PN","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_16","alias_value":"ZPXHUXBLB6PN77NK","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_8","alias_value":"ZPXHUXBL","created_at":"2026-07-05T04:08:33Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2022:ZPXHUXBLB6PN77NKG5JK4KIIXB","target":"record","payload":{"canonical_record":{"source":{"id":"2203.13553","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-25T10:19:35Z","cross_cats_sorted":[],"title_canon_sha256":"c9c6852b9ee495912ef8300870d3438748f1c336e2bc3d0caac0c67983974fff","abstract_canon_sha256":"7685008200a6d15d446e56e1f2f32842022cda30fd687b5be99e862e2e90df42"},"schema_version":"1.0"},"canonical_sha256":"cbee7a5c2b0f9edffdaa3752ae2908b852b2bf60038c06737493d8cd3c9998ca","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T04:08:33.351783Z","signature_b64":"lMdwhNDd+3eCRQ/wakSxvSJ8ZtihHOUk/sDnhDBWUTiCt/V5sEGr5nEaL8SUYinqYn1gY++IexC5jfCOhXCpAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"cbee7a5c2b0f9edffdaa3752ae2908b852b2bf60038c06737493d8cd3c9998ca","last_reissued_at":"2026-07-05T04:08:33.351334Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T04:08:33.351334Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2203.13553","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":"2026-07-05T04:08:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3ctFfsSzz0O/7pyeNTNyDrFDmbcw0y3jO66vvi7SWfV5EDQufIvI5O3BHR6bA+jo8VUnDZocbp8QUood8XWhBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T01:44:31.306791Z"},"content_sha256":"05c9aefbc925f80185d537b7a46b2ecb5ef4c5145049eab88428278c33d5a1f5","schema_version":"1.0","event_id":"sha256:05c9aefbc925f80185d537b7a46b2ecb5ef4c5145049eab88428278c33d5a1f5"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2022:ZPXHUXBLB6PN77NKG5JK4KIIXB","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Preprocessing Reward Functions for Interpretability","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Adam Gleave, Erik Jenner","submitted_at":"2022-03-25T10:19:35Z","abstract_excerpt":"In many real-world applications, the reward function is too complex to be manually specified. In such cases, reward functions must instead be learned from human feedback. Since the learned reward may fail to represent user preferences, it is important to be able to validate the learned reward function prior to deployment. One promising approach is to apply interpretability tools to the reward function to spot potential deviations from the user's intention. Existing work has applied general-purpose interpretability tools to understand learned reward functions. We propose exploiting the intrinsi"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.13553","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/2203.13553/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":"2026-07-05T04:08:33Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"co4NFdewpBznRw6DKmIE6fsY1b1vtn2i/IMavsUOlzKW880YVRuk0KLig8aJzrtuGpM66tQEtl54kQOlOJFQAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-17T01:44:31.307312Z"},"content_sha256":"54aa0c672f764ab959275c3c07fd2cf8f1e7a5b596ad67db67625d377d782e4f","schema_version":"1.0","event_id":"sha256:54aa0c672f764ab959275c3c07fd2cf8f1e7a5b596ad67db67625d377d782e4f"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/bundle.json","state_url":"https://pith.science/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/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":"2026-08-17T01:44:31Z","links":{"resolver":"https://pith.science/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB","bundle":"https://pith.science/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/bundle.json","state":"https://pith.science/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/state.json","well_known_bundle":"https://pith.science/.well-known/pith/ZPXHUXBLB6PN77NKG5JK4KIIXB/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2022:ZPXHUXBLB6PN77NKG5JK4KIIXB","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":"7685008200a6d15d446e56e1f2f32842022cda30fd687b5be99e862e2e90df42","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-25T10:19:35Z","title_canon_sha256":"c9c6852b9ee495912ef8300870d3438748f1c336e2bc3d0caac0c67983974fff"},"schema_version":"1.0","source":{"id":"2203.13553","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2203.13553","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"arxiv_version","alias_value":"2203.13553v1","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2203.13553","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_12","alias_value":"ZPXHUXBLB6PN","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_16","alias_value":"ZPXHUXBLB6PN77NK","created_at":"2026-07-05T04:08:33Z"},{"alias_kind":"pith_short_8","alias_value":"ZPXHUXBL","created_at":"2026-07-05T04:08:33Z"}],"graph_snapshots":[{"event_id":"sha256:54aa0c672f764ab959275c3c07fd2cf8f1e7a5b596ad67db67625d377d782e4f","target":"graph","created_at":"2026-07-05T04:08:33Z","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/2203.13553/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"In many real-world applications, the reward function is too complex to be manually specified. In such cases, reward functions must instead be learned from human feedback. Since the learned reward may fail to represent user preferences, it is important to be able to validate the learned reward function prior to deployment. One promising approach is to apply interpretability tools to the reward function to spot potential deviations from the user's intention. Existing work has applied general-purpose interpretability tools to understand learned reward functions. We propose exploiting the intrinsi","authors_text":"Adam Gleave, Erik Jenner","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-25T10:19:35Z","title":"Preprocessing Reward Functions for Interpretability"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2203.13553","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":"sha256:05c9aefbc925f80185d537b7a46b2ecb5ef4c5145049eab88428278c33d5a1f5","target":"record","created_at":"2026-07-05T04:08:33Z","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":"7685008200a6d15d446e56e1f2f32842022cda30fd687b5be99e862e2e90df42","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2022-03-25T10:19:35Z","title_canon_sha256":"c9c6852b9ee495912ef8300870d3438748f1c336e2bc3d0caac0c67983974fff"},"schema_version":"1.0","source":{"id":"2203.13553","kind":"arxiv","version":1}},"canonical_sha256":"cbee7a5c2b0f9edffdaa3752ae2908b852b2bf60038c06737493d8cd3c9998ca","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"cbee7a5c2b0f9edffdaa3752ae2908b852b2bf60038c06737493d8cd3c9998ca","first_computed_at":"2026-07-05T04:08:33.351334Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T04:08:33.351334Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"lMdwhNDd+3eCRQ/wakSxvSJ8ZtihHOUk/sDnhDBWUTiCt/V5sEGr5nEaL8SUYinqYn1gY++IexC5jfCOhXCpAg==","signature_status":"signed_v1","signed_at":"2026-07-05T04:08:33.351783Z","signed_message":"canonical_sha256_bytes"},"source_id":"2203.13553","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:05c9aefbc925f80185d537b7a46b2ecb5ef4c5145049eab88428278c33d5a1f5","sha256:54aa0c672f764ab959275c3c07fd2cf8f1e7a5b596ad67db67625d377d782e4f"],"state_sha256":"7e99c6b223266fa2bf89cae32afaf14ee8966755d95403546fd3e14662b55999"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"af6nCRsx7WS4ePE7Lj6y0wwaIcsTTSyGs7xFW+/H03yiqwNMRCWtcC9aUbDWFS2Xi3rLVbPQiGAVppAyN3OgDA==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-17T01:44:31.311664Z","bundle_sha256":"d18667ae3a1625c8e375e64d8d49083b9c82738e106948b89486f17527f6d586"}}