{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","short_pith_number":"pith:S4KIC3SA","canonical_record":{"source":{"id":"2607.04590","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-06T01:33:12Z","cross_cats_sorted":[],"title_canon_sha256":"7fb9fbab7e65e0e2cb47a9ab233b47a7828ff25d130708f8add853f83339dcfc","abstract_canon_sha256":"2e4e2a0d767f1ced054029b33077b3acc55c7cc71ba264bf1b9d92cac90a11b6"},"schema_version":"1.0"},"canonical_sha256":"9714816e408518823d9f42aa3af6dfb283493faccce1439f100738503887d862","source":{"kind":"arxiv","id":"2607.04590","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.04590","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"arxiv_version","alias_value":"2607.04590v1","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.04590","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_12","alias_value":"S4KIC3SAQUMI","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_16","alias_value":"S4KIC3SAQUMIEPM7","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_8","alias_value":"S4KIC3SA","created_at":"2026-07-07T02:19:26Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"record","payload":{"canonical_record":{"source":{"id":"2607.04590","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-06T01:33:12Z","cross_cats_sorted":[],"title_canon_sha256":"7fb9fbab7e65e0e2cb47a9ab233b47a7828ff25d130708f8add853f83339dcfc","abstract_canon_sha256":"2e4e2a0d767f1ced054029b33077b3acc55c7cc71ba264bf1b9d92cac90a11b6"},"schema_version":"1.0"},"canonical_sha256":"9714816e408518823d9f42aa3af6dfb283493faccce1439f100738503887d862","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-07T02:19:26.095732Z","signature_b64":"fiWPE4jLfhYxrA8l2/bZVv8M2fdGUnZv5Xc7sxnFIH3FRGmgcw091RqUVlMgHJI/izFioFMy/1QsBnFiKCvFDQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"9714816e408518823d9f42aa3af6dfb283493faccce1439f100738503887d862","last_reissued_at":"2026-07-07T02:19:26.094896Z","signature_status":"signed_v1","first_computed_at":"2026-07-07T02:19:26.094896Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.04590","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":"2026-07-07T02:19:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"A0sFUjhXKCgK/B0E4CDD+DNDcGn5iiZu3xm0NwYBtahDWKfH+yhR08Mfu7zQdazO54kcY9KGo5ODa0wJhoEpAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.344784Z"},"content_sha256":"5570f2291fe15f430cd1007d54a027c8c24559069d5219fb86afe088d4a3cb2b","schema_version":"1.0","event_id":"sha256:5570f2291fe15f430cd1007d54a027c8c24559069d5219fb86afe088d4a3cb2b"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Attention Limited Reward Learning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Wenqian Xing","submitted_at":"2026-07-06T01:33:12Z","abstract_excerpt":"Pairwise human comparisons are a primary interface through which modern AI systems learn human preferences. RLHF and related alignment pipelines typically model such comparisons with Bradley--Terry log-odds, where choice probabilities are governed by latent reward differences. This paper examines what this assumption misses through a reduced-form model motivated by rational inattention, in which each label is generated by a low-capacity evaluation channel. The model separates two forms of ambiguity that standard reward modeling tends to conflate: a comparison may be difficult because the two c"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.04590","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/2607.04590/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":"2026-07-07T02:19:26Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"IdDXEn1LBCRW58dbok7sVn8touppwC1IaXe/0MPR7LdGML0bVHLikmqw8TwKau/4lnjYfzMtmUXRKDPfnf49Ag==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.345881Z"},"content_sha256":"7c5b1e57e79e185b317135362b65505a39aee66a6517349525248bbdf08eaf02","schema_version":"1.0","event_id":"sha256:7c5b1e57e79e185b317135362b65505a39aee66a6517349525248bbdf08eaf02"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.3389/fnins.2017.00468/full.18) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Gabriela Tavares, Pietro Perona, and Antonio Rangel. The attentional drift diffusion model of simple perceptual decision-making.Frontiers in Neuroscience, 11:468, 2017. doi: 10.3389/ fnins.2017.00468. URL https://www.frontiersin.org/journal","arxiv_id":"2607.04590","detector":"doi_compliance","evidence":{"ref_index":35,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Gabriela Tavares, Pietro Perona, and Antonio Rangel. The attentional drift diffusion model of simple perceptual decision-making.Frontiers in Neuroscience, 11:468, 2017. doi: 10.3389/ fnins.2017.00468. URL https://www.frontiersin.org/journal","reconstructed_doi":"10.3389/fnins.2017.00468/full.18"},"severity":"advisory","ref_index":35,"audited_at":"2026-07-11T16:51:38.447715Z","event_type":"pith.integrity.v1","detected_doi":"10.3389/fnins.2017.00468/full.18","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"d234a5b3d4befd848f0c581461b69529b47e7e0c151fe8b7d32b9336c1f90e03","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":12232,"payload_sha256":"4afd090a733883aa83517e5690aa52c436e7e23c2a35a6f3bf3e6503ee1666c2","signature_b64":"b6ErQ8/aO97okKUoW++KK8/zt6xohwojo7TXKzpC7VrDbT4hh42rBPHoer/NYW9dlyvXqPWKfJeFmME7e7c4Dw==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-11T16:56:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"HXQltPMvEaLA299C9dFwcPtVlsatNaVqsLu2JRNYHbD9OaY7T15sizPCETfF1yGNDvbxntDDV3jGrQpx4IF/Aw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.435091Z"},"content_sha256":"655a3b1d7b61abf60952cb0b9802233cc21019ea962b3833789d17631c70792d","schema_version":"1.0","event_id":"sha256:655a3b1d7b61abf60952cb0b9802233cc21019ea962b3833789d17631c70792d"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.48550/arXiv.2209.13085.URL) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Joar Skalse, Nikolaus Howe, Dmitrii Krasheninnikov, and David Krueger. Defin- ing and characterizing reward gaming. InAdvances in Neural Information Pro- cessing Systems, volume 35, pages 9460–9471, 2022. doi: 10.48550/arXiv. 2209.13085. UR","arxiv_id":"2607.04590","detector":"doi_compliance","evidence":{"ref_index":32,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Joar Skalse, Nikolaus Howe, Dmitrii Krasheninnikov, and David Krueger. Defin- ing and characterizing reward gaming. InAdvances in Neural Information Pro- cessing Systems, volume 35, pages 9460–9471, 2022. doi: 10.48550/arXiv. 2209.13085. UR","reconstructed_doi":"10.48550/arXiv.2209.13085.URL"},"severity":"advisory","ref_index":32,"audited_at":"2026-07-11T16:51:38.447715Z","event_type":"pith.integrity.v1","detected_doi":"10.48550/arXiv.2209.13085.URL","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"eda834ca35cea9b1a5e2d1f28519558fa604fa340061c7f0f2a8a5c90eb87900","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":12231,"payload_sha256":"30f9153476302a87ed4d617e814a4a3d248fc77ccfa52e2aeacb63d447ae6953","signature_b64":"vTBt4qAWVCV/wLR6KCngiYAsLMmWkEqGNa9bg8f3MFsy7AK05zjXEKU60bC6/qmLPtQ6NidxZ1bKSOu9KUywCw==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-11T16:56:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"3Xu+p4N7ZW2Yyi+3urma9Jctau2LreW675JMsOBBeeXd8OgV8hbltZ31Uzpllq19XvK7TixyEAxJZYMVs7kiBQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.435736Z"},"content_sha256":"dc6b5f05724581891ca40117866bdb9f50248e2a20205c450543dd2cfc8c2646","schema_version":"1.0","event_id":"sha256:dc6b5f05724581891ca40117866bdb9f50248e2a20205c450543dd2cfc8c2646"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.15607/RSS.2017.XIII.053.URLhttps://www.roboticsproceedings.org/rss13/p53.html) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Dorsa Sadigh, Anca D. Dragan, S. Shankar Sastry, and Sanjit A. Seshia. Active preference-based learning of reward functions. InRobotics: Science and Systems XIII, 2017. doi: 10.15607/RSS. 2017.XIII.053. URLhttps://www.roboticsproceedings.or","arxiv_id":"2607.04590","detector":"doi_compliance","evidence":{"ref_index":26,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Dorsa Sadigh, Anca D. Dragan, S. Shankar Sastry, and Sanjit A. Seshia. Active preference-based learning of reward functions. InRobotics: Science and Systems XIII, 2017. doi: 10.15607/RSS. 2017.XIII.053. URLhttps://www.roboticsproceedings.or","reconstructed_doi":"10.15607/RSS.2017.XIII.053.URLhttps://www.roboticsproceedings.org/rss13/p53.html"},"severity":"advisory","ref_index":26,"audited_at":"2026-07-11T16:51:38.447715Z","event_type":"pith.integrity.v1","detected_doi":"10.15607/RSS.2017.XIII.053.URLhttps://www.roboticsproceedings.org/rss13/p53.html","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"adbcd347e8aa0f58e8eef59fde686e6b145589ef8c9e2b7c871f5f623ecb0667","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":12230,"payload_sha256":"ceb699baabea1d23c8f847fd0756a15916471fd7bbc64ac7dae566b6cc686ef2","signature_b64":"yITRsG4JKCLoHF5JInD0UmDpf9X4bVQnJH/sbQN+FMHcMZ5V3ekQ9dRTLFL6sn6aFH8V9pEDdgqiZqeyRC4HDA==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-11T16:56:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"yP/FPOUaOW8/ySDr90V1xTOVr6AhyH+4gBRK+aowSiXsWoGpi3e9ZFEawsHCJWQZOA9nBrdlbJAEkRgLlVZnBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.436459Z"},"content_sha256":"8bdb76c19ecf846e4516437f6871390d18e442e2c9f5b807aec89393f690cce1","schema_version":"1.0","event_id":"sha256:8bdb76c19ecf846e4516437f6871390d18e442e2c9f5b807aec89393f690cce1"},{"event_type":"integrity_finding","subject_pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","target":"integrity","payload":{"note":"DOI in the printed bibliography is fragmented by whitespace or line breaks. A longer candidate (10.1093/biomet/39.3-4.324.15) was visible in the surrounding text but could not be confirmed against doi.org as printed.","snippet":"Ralph Allan Bradley and Milton E. Terry. Rank analysis of incomplete block designs: I. the method of paired comparisons.Biometrika, 39(3/4):324–345, 1952. doi: 10.1093/biomet/39. 3-4.324. 15","arxiv_id":"2607.04590","detector":"doi_compliance","evidence":{"ref_index":3,"verdict_class":"incontrovertible","resolved_title":null,"printed_excerpt":"Ralph Allan Bradley and Milton E. Terry. Rank analysis of incomplete block designs: I. the method of paired comparisons.Biometrika, 39(3/4):324–345, 1952. doi: 10.1093/biomet/39. 3-4.324. 15","reconstructed_doi":"10.1093/biomet/39.3-4.324.15"},"severity":"advisory","ref_index":3,"audited_at":"2026-07-11T16:51:38.447715Z","event_type":"pith.integrity.v1","detected_doi":"10.1093/biomet/39.3-4.324.15","detector_url":"https://pith.science/pith-integrity-protocol#doi_compliance","external_url":null,"finding_type":"recoverable_identifier","evidence_hash":"f19f3fd704d7f2127593d16b258bf9307d86c6a3d9f9639d55ad9e4001d86108","paper_version":1,"verdict_class":"incontrovertible","resolved_title":null,"detector_version":"1.0.0","detected_arxiv_id":null,"integrity_event_id":12229,"payload_sha256":"e54c4121cee240b12c55160ae99f821d005955c67a69df4164b6e1d3946d0359","signature_b64":"TFTvXz+L+4XRaf73mH7mpri0kc/1aregqdEsaRO8aw0RLSuu90KTl09y5qq/pUEgAUjvKCj4bEXpVYl05c0VCA==","signing_key_id":"pith-v1-2026-05"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-11T16:56:23Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"bijLa0y27CasYhJ1ymmsaGTl8Qa9tv78X52CYt527Wgi5bwkmrFR5sqaz+Cfx3R2RQcgoKD78b6JfBbqA5YYAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-06T09:23:35.437255Z"},"content_sha256":"b6694713bdb8bbc21b3cfd4ef51ca391f885be4fad533b94671e872976532bd3","schema_version":"1.0","event_id":"sha256:b6694713bdb8bbc21b3cfd4ef51ca391f885be4fad533b94671e872976532bd3"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/bundle.json","state_url":"https://pith.science/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/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":"2026-08-06T09:23:35Z","links":{"resolver":"https://pith.science/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK","bundle":"https://pith.science/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/bundle.json","state":"https://pith.science/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/state.json","well_known_bundle":"https://pith.science/.well-known/pith/S4KIC3SAQUMIEPM7IKVDV5W7WK/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:S4KIC3SAQUMIEPM7IKVDV5W7WK","merge_version":"pith-open-graph-merge-v1","event_count":6,"valid_event_count":6,"invalid_event_count":0,"equivocation_count":1,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"2e4e2a0d767f1ced054029b33077b3acc55c7cc71ba264bf1b9d92cac90a11b6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-06T01:33:12Z","title_canon_sha256":"7fb9fbab7e65e0e2cb47a9ab233b47a7828ff25d130708f8add853f83339dcfc"},"schema_version":"1.0","source":{"id":"2607.04590","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.04590","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"arxiv_version","alias_value":"2607.04590v1","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.04590","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_12","alias_value":"S4KIC3SAQUMI","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_16","alias_value":"S4KIC3SAQUMIEPM7","created_at":"2026-07-07T02:19:26Z"},{"alias_kind":"pith_short_8","alias_value":"S4KIC3SA","created_at":"2026-07-07T02:19:26Z"}],"graph_snapshots":[{"event_id":"sha256:7c5b1e57e79e185b317135362b65505a39aee66a6517349525248bbdf08eaf02","target":"graph","created_at":"2026-07-07T02:19:26Z","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/2607.04590/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Pairwise human comparisons are a primary interface through which modern AI systems learn human preferences. RLHF and related alignment pipelines typically model such comparisons with Bradley--Terry log-odds, where choice probabilities are governed by latent reward differences. This paper examines what this assumption misses through a reduced-form model motivated by rational inattention, in which each label is generated by a low-capacity evaluation channel. The model separates two forms of ambiguity that standard reward modeling tends to conflate: a comparison may be difficult because the two c","authors_text":"Wenqian Xing","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-06T01:33:12Z","title":"Attention Limited Reward Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.04590","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":"sha256:5570f2291fe15f430cd1007d54a027c8c24559069d5219fb86afe088d4a3cb2b","target":"record","created_at":"2026-07-07T02:19:26Z","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":"2e4e2a0d767f1ced054029b33077b3acc55c7cc71ba264bf1b9d92cac90a11b6","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-06T01:33:12Z","title_canon_sha256":"7fb9fbab7e65e0e2cb47a9ab233b47a7828ff25d130708f8add853f83339dcfc"},"schema_version":"1.0","source":{"id":"2607.04590","kind":"arxiv","version":1}},"canonical_sha256":"9714816e408518823d9f42aa3af6dfb283493faccce1439f100738503887d862","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"9714816e408518823d9f42aa3af6dfb283493faccce1439f100738503887d862","first_computed_at":"2026-07-07T02:19:26.094896Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-07T02:19:26.094896Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"fiWPE4jLfhYxrA8l2/bZVv8M2fdGUnZv5Xc7sxnFIH3FRGmgcw091RqUVlMgHJI/izFioFMy/1QsBnFiKCvFDQ==","signature_status":"signed_v1","signed_at":"2026-07-07T02:19:26.095732Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.04590","source_kind":"arxiv","source_version":1}}},"equivocations":[{"signer_id":"pith.science","event_type":"integrity_finding","target":"integrity","event_ids":["sha256:655a3b1d7b61abf60952cb0b9802233cc21019ea962b3833789d17631c70792d","sha256:8bdb76c19ecf846e4516437f6871390d18e442e2c9f5b807aec89393f690cce1","sha256:b6694713bdb8bbc21b3cfd4ef51ca391f885be4fad533b94671e872976532bd3","sha256:dc6b5f05724581891ca40117866bdb9f50248e2a20205c450543dd2cfc8c2646"]}],"invalid_events":[],"applied_event_ids":["sha256:5570f2291fe15f430cd1007d54a027c8c24559069d5219fb86afe088d4a3cb2b","sha256:7c5b1e57e79e185b317135362b65505a39aee66a6517349525248bbdf08eaf02"],"state_sha256":"fd9f366a3f76253f563a60d13bedd14fb96dba4a80f92f5ef7c57125fe06fb71"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"c8tTzsEngEsO2G2gHC4K9fdGqwamzrYGrQLg4aggQLklJ2UMxyItqgeHtJFy0mRZKo2VO+5a2esPsNAcSvr5BQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-06T09:23:35.444598Z","bundle_sha256":"01fcd5fb7bbfd9d43cea744c56627dc79851aa88a5c5e083360a310de47394c7"}}