{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2026:K4QC7AD37YPFYKXY4NXGJM5SJA","short_pith_number":"pith:K4QC7AD3","canonical_record":{"source":{"id":"2607.09786","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T14:18:26Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"09cd21e1830a144a1aeb25c5948bde91d336affff8267ded711c6c48dea6f53b","abstract_canon_sha256":"f66005ade48ca68546466db9c26c2af851025c594e6f64841ee38f13a12ed859"},"schema_version":"1.0"},"canonical_sha256":"57202f807bfe1e5c2af8e36e64b3b248080166f1d2b43f5d1e39b7547790657d","source":{"kind":"arxiv","id":"2607.09786","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.09786","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2607.09786v1","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.09786","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"K4QC7AD37YPF","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"K4QC7AD37YPFYKXY","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"K4QC7AD3","created_at":"2026-07-14T00:18:35Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2026:K4QC7AD37YPFYKXY4NXGJM5SJA","target":"record","payload":{"canonical_record":{"source":{"id":"2607.09786","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T14:18:26Z","cross_cats_sorted":["cs.CL","cs.LG"],"title_canon_sha256":"09cd21e1830a144a1aeb25c5948bde91d336affff8267ded711c6c48dea6f53b","abstract_canon_sha256":"f66005ade48ca68546466db9c26c2af851025c594e6f64841ee38f13a12ed859"},"schema_version":"1.0"},"canonical_sha256":"57202f807bfe1e5c2af8e36e64b3b248080166f1d2b43f5d1e39b7547790657d","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-14T00:18:35.697939Z","signature_b64":"Q7bYp/ULFfE55oWfFmEHxFmqPUl03eRMk9l3kquXHYUumUS+4CtYVHDHCtBjco/9Rh47xC/P1FKGka3x3P8yBg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"57202f807bfe1e5c2af8e36e64b3b248080166f1d2b43f5d1e39b7547790657d","last_reissued_at":"2026-07-14T00:18:35.697036Z","signature_status":"signed_v1","first_computed_at":"2026-07-14T00:18:35.697036Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2607.09786","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":"2026-07-14T00:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"n2k9QhxyX5i3AW+aojCiyN48LqFHQYuVvh3wXqJM2jcc85jUECG30rjM/zLZYCq849+lgn4hURtg5PbsI0pABQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:48:42.420290Z"},"content_sha256":"53de03dbfb4143cd6e6b8d5292e6a72dc595fee9f5ccfa661bdb8ba077a9664a","schema_version":"1.0","event_id":"sha256:53de03dbfb4143cd6e6b8d5292e6a72dc595fee9f5ccfa661bdb8ba077a9664a"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2026:K4QC7AD37YPFYKXY4NXGJM5SJA","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Length Penalties Make Chain-of-Thought Less Monitorable","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL","cs.LG"],"primary_cat":"cs.AI","authors_text":"Bryce Little","submitted_at":"2026-07-08T14:18:26Z","abstract_excerpt":"Length-penalized reinforcement learning can shorten chain-of-thought reasoning while hiding an influence that drives the model's answer. In our experiments, training with length penalties does not stop misleading hints from steering models, even though the models' chains of thought mention the hint much less often. A token-accuracy evaluation would count these runs as successful because they use fewer reasoning tokens with little accuracy loss; it would miss whether the remaining trace still shows what drove the answer. We train Qwen3-4B and Qwen3-14B variants with different target chain lengt"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.09786","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/2607.09786/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":"2026-07-14T00:18:35Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"xf8dXjG2MLIZHZW8IcJT1eoZM9oOpHZlf/somfgoTsU16Kwf5mB/tf3BnM8WpF0QhHuNOqCti4xxGmk4h/ahCQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-03T22:48:42.420832Z"},"content_sha256":"bfa36426cc7f636fd58ccc3cc3025dccf9666f341f49582ad347350eee099a96","schema_version":"1.0","event_id":"sha256:bfa36426cc7f636fd58ccc3cc3025dccf9666f341f49582ad347350eee099a96"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/bundle.json","state_url":"https://pith.science/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/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":"2026-08-03T22:48:42Z","links":{"resolver":"https://pith.science/pith/K4QC7AD37YPFYKXY4NXGJM5SJA","bundle":"https://pith.science/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/bundle.json","state":"https://pith.science/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/state.json","well_known_bundle":"https://pith.science/.well-known/pith/K4QC7AD37YPFYKXY4NXGJM5SJA/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2026:K4QC7AD37YPFYKXY4NXGJM5SJA","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":"f66005ade48ca68546466db9c26c2af851025c594e6f64841ee38f13a12ed859","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T14:18:26Z","title_canon_sha256":"09cd21e1830a144a1aeb25c5948bde91d336affff8267ded711c6c48dea6f53b"},"schema_version":"1.0","source":{"id":"2607.09786","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2607.09786","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"arxiv_version","alias_value":"2607.09786v1","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2607.09786","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_12","alias_value":"K4QC7AD37YPF","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_16","alias_value":"K4QC7AD37YPFYKXY","created_at":"2026-07-14T00:18:35Z"},{"alias_kind":"pith_short_8","alias_value":"K4QC7AD3","created_at":"2026-07-14T00:18:35Z"}],"graph_snapshots":[{"event_id":"sha256:bfa36426cc7f636fd58ccc3cc3025dccf9666f341f49582ad347350eee099a96","target":"graph","created_at":"2026-07-14T00:18:35Z","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/2607.09786/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Length-penalized reinforcement learning can shorten chain-of-thought reasoning while hiding an influence that drives the model's answer. In our experiments, training with length penalties does not stop misleading hints from steering models, even though the models' chains of thought mention the hint much less often. A token-accuracy evaluation would count these runs as successful because they use fewer reasoning tokens with little accuracy loss; it would miss whether the remaining trace still shows what drove the answer. We train Qwen3-4B and Qwen3-14B variants with different target chain lengt","authors_text":"Bryce Little","cross_cats":["cs.CL","cs.LG"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T14:18:26Z","title":"Length Penalties Make Chain-of-Thought Less Monitorable"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2607.09786","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":"sha256:53de03dbfb4143cd6e6b8d5292e6a72dc595fee9f5ccfa661bdb8ba077a9664a","target":"record","created_at":"2026-07-14T00:18:35Z","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":"f66005ade48ca68546466db9c26c2af851025c594e6f64841ee38f13a12ed859","cross_cats_sorted":["cs.CL","cs.LG"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2026-07-08T14:18:26Z","title_canon_sha256":"09cd21e1830a144a1aeb25c5948bde91d336affff8267ded711c6c48dea6f53b"},"schema_version":"1.0","source":{"id":"2607.09786","kind":"arxiv","version":1}},"canonical_sha256":"57202f807bfe1e5c2af8e36e64b3b248080166f1d2b43f5d1e39b7547790657d","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"57202f807bfe1e5c2af8e36e64b3b248080166f1d2b43f5d1e39b7547790657d","first_computed_at":"2026-07-14T00:18:35.697036Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-14T00:18:35.697036Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Q7bYp/ULFfE55oWfFmEHxFmqPUl03eRMk9l3kquXHYUumUS+4CtYVHDHCtBjco/9Rh47xC/P1FKGka3x3P8yBg==","signature_status":"signed_v1","signed_at":"2026-07-14T00:18:35.697939Z","signed_message":"canonical_sha256_bytes"},"source_id":"2607.09786","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:53de03dbfb4143cd6e6b8d5292e6a72dc595fee9f5ccfa661bdb8ba077a9664a","sha256:bfa36426cc7f636fd58ccc3cc3025dccf9666f341f49582ad347350eee099a96"],"state_sha256":"4a8a4bb5e4f72ab8cf34965cc24d1118e7a512c23a2455a2b14126c1506b48a4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"QmjKao8zeibFZj0A6rSu6RWII6qH5yWupYc+Pn5sioTwtDLGmd0/yhGJ+nzaid0YaQ1m32MlDJtIvNOH6us3DQ==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-03T22:48:42.426981Z","bundle_sha256":"a9ac855a926e92e8376e197731791762467803337c4da56c45b69b2785ac74db"}}