{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2016:OTCIM2WH27COCLBA6EIB7FGOYE","short_pith_number":"pith:OTCIM2WH","canonical_record":{"source":{"id":"1605.03143","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-05-10T18:28:57Z","cross_cats_sorted":[],"title_canon_sha256":"e6cc9ead74d3467ccac97be7e771faf8fd831a05966671e53ba63df7aad24728","abstract_canon_sha256":"0f299bacaf603ec01594c49da7691b7d61c34222f66265e451ac6eea3d2f9086"},"schema_version":"1.0"},"canonical_sha256":"74c4866ac7d7c4e12c20f1101f94cec11e43f31998ae3fa95963fdaab4325750","source":{"kind":"arxiv","id":"1605.03143","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1605.03143","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"arxiv_version","alias_value":"1605.03143v1","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1605.03143","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"pith_short_12","alias_value":"OTCIM2WH27CO","created_at":"2026-05-18T12:30:36Z"},{"alias_kind":"pith_short_16","alias_value":"OTCIM2WH27COCLBA","created_at":"2026-05-18T12:30:36Z"},{"alias_kind":"pith_short_8","alias_value":"OTCIM2WH","created_at":"2026-05-18T12:30:36Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2016:OTCIM2WH27COCLBA6EIB7FGOYE","target":"record","payload":{"canonical_record":{"source":{"id":"1605.03143","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-05-10T18:28:57Z","cross_cats_sorted":[],"title_canon_sha256":"e6cc9ead74d3467ccac97be7e771faf8fd831a05966671e53ba63df7aad24728","abstract_canon_sha256":"0f299bacaf603ec01594c49da7691b7d61c34222f66265e451ac6eea3d2f9086"},"schema_version":"1.0"},"canonical_sha256":"74c4866ac7d7c4e12c20f1101f94cec11e43f31998ae3fa95963fdaab4325750","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-18T01:15:07.436666Z","signature_b64":"549GyB07SlzBeo/Wm1KE3bNKCxNSFIgatmOxLCu6v2PuBNpI6vfq6TlysNR+N7s4xNfkmRoxzNtIDsJJNyGaAA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"74c4866ac7d7c4e12c20f1101f94cec11e43f31998ae3fa95963fdaab4325750","last_reissued_at":"2026-05-18T01:15:07.435920Z","signature_status":"signed_v1","first_computed_at":"2026-05-18T01:15:07.435920Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"1605.03143","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":"2026-05-18T01:15:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"Uh5zNtwbCXm1DKpdZNEGGxR4ivhBkVX7zcG080tF/7nZGSGV7SMPDfx107VWT/tAfi69/ftfy55Ev0Sshpf7BQ==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T13:36:29.181364Z"},"content_sha256":"46596a656a99e3a9654815b2cebc082d98685b13cedb0fe1742da322329f0aec","schema_version":"1.0","event_id":"sha256:46596a656a99e3a9654815b2cebc082d98685b13cedb0fe1742da322329f0aec"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2016:OTCIM2WH27COCLBA6EIB7FGOYE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Avoiding Wireheading with Value Reinforcement Learning","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Marcus Hutter, Tom Everitt","submitted_at":"2016-05-10T18:28:57Z","abstract_excerpt":"How can we design good goals for arbitrarily intelligent agents? Reinforcement learning (RL) is a natural approach. Unfortunately, RL does not work well for generally intelligent agents, as RL agents are incentivised to shortcut the reward sensor for maximum reward -- the so-called wireheading problem. In this paper we suggest an alternative to RL called value reinforcement learning (VRL). In VRL, agents use the reward signal to learn a utility function. The VRL setup allows us to remove the incentive to wirehead by placing a constraint on the agent's actions. The constraint is defined in term"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1605.03143","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":""},"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":"2026-05-18T01:15:07Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"PFxv9r/VPnT721y51AAml5KaExFXrmBHWYOUXigkRJWETQDBXBTiIHiE5M/NBz2X6ZeLeKX+NBHR56RvFUxQDg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-06-05T13:36:29.181717Z"},"content_sha256":"0bcafb6d4c44069feac0e38469b127d9cdd62f6be58336b08046d1ce36958373","schema_version":"1.0","event_id":"sha256:0bcafb6d4c44069feac0e38469b127d9cdd62f6be58336b08046d1ce36958373"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/OTCIM2WH27COCLBA6EIB7FGOYE/bundle.json","state_url":"https://pith.science/pith/OTCIM2WH27COCLBA6EIB7FGOYE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/OTCIM2WH27COCLBA6EIB7FGOYE/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":"2026-06-05T13:36:29Z","links":{"resolver":"https://pith.science/pith/OTCIM2WH27COCLBA6EIB7FGOYE","bundle":"https://pith.science/pith/OTCIM2WH27COCLBA6EIB7FGOYE/bundle.json","state":"https://pith.science/pith/OTCIM2WH27COCLBA6EIB7FGOYE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/OTCIM2WH27COCLBA6EIB7FGOYE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2016:OTCIM2WH27COCLBA6EIB7FGOYE","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":"0f299bacaf603ec01594c49da7691b7d61c34222f66265e451ac6eea3d2f9086","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-05-10T18:28:57Z","title_canon_sha256":"e6cc9ead74d3467ccac97be7e771faf8fd831a05966671e53ba63df7aad24728"},"schema_version":"1.0","source":{"id":"1605.03143","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"1605.03143","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"arxiv_version","alias_value":"1605.03143v1","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.1605.03143","created_at":"2026-05-18T01:15:07Z"},{"alias_kind":"pith_short_12","alias_value":"OTCIM2WH27CO","created_at":"2026-05-18T12:30:36Z"},{"alias_kind":"pith_short_16","alias_value":"OTCIM2WH27COCLBA","created_at":"2026-05-18T12:30:36Z"},{"alias_kind":"pith_short_8","alias_value":"OTCIM2WH","created_at":"2026-05-18T12:30:36Z"}],"graph_snapshots":[{"event_id":"sha256:0bcafb6d4c44069feac0e38469b127d9cdd62f6be58336b08046d1ce36958373","target":"graph","created_at":"2026-05-18T01:15:07Z","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"},"paper":{"abstract_excerpt":"How can we design good goals for arbitrarily intelligent agents? Reinforcement learning (RL) is a natural approach. Unfortunately, RL does not work well for generally intelligent agents, as RL agents are incentivised to shortcut the reward sensor for maximum reward -- the so-called wireheading problem. In this paper we suggest an alternative to RL called value reinforcement learning (VRL). In VRL, agents use the reward signal to learn a utility function. The VRL setup allows us to remove the incentive to wirehead by placing a constraint on the agent's actions. The constraint is defined in term","authors_text":"Marcus Hutter, Tom Everitt","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-05-10T18:28:57Z","title":"Avoiding Wireheading with Value Reinforcement Learning"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"1605.03143","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":"sha256:46596a656a99e3a9654815b2cebc082d98685b13cedb0fe1742da322329f0aec","target":"record","created_at":"2026-05-18T01:15:07Z","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":"0f299bacaf603ec01594c49da7691b7d61c34222f66265e451ac6eea3d2f9086","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.AI","submitted_at":"2016-05-10T18:28:57Z","title_canon_sha256":"e6cc9ead74d3467ccac97be7e771faf8fd831a05966671e53ba63df7aad24728"},"schema_version":"1.0","source":{"id":"1605.03143","kind":"arxiv","version":1}},"canonical_sha256":"74c4866ac7d7c4e12c20f1101f94cec11e43f31998ae3fa95963fdaab4325750","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"74c4866ac7d7c4e12c20f1101f94cec11e43f31998ae3fa95963fdaab4325750","first_computed_at":"2026-05-18T01:15:07.435920Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-05-18T01:15:07.435920Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"549GyB07SlzBeo/Wm1KE3bNKCxNSFIgatmOxLCu6v2PuBNpI6vfq6TlysNR+N7s4xNfkmRoxzNtIDsJJNyGaAA==","signature_status":"signed_v1","signed_at":"2026-05-18T01:15:07.436666Z","signed_message":"canonical_sha256_bytes"},"source_id":"1605.03143","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:46596a656a99e3a9654815b2cebc082d98685b13cedb0fe1742da322329f0aec","sha256:0bcafb6d4c44069feac0e38469b127d9cdd62f6be58336b08046d1ce36958373"],"state_sha256":"1208a0548a38fd52601c06360c06e8be22d13ffe5294e93389b8756ca79050f4"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"gdzG+gOi4td5N1YDl5+YObq4ZS+p1kb4vbTZ4TRwc4O6k0upVmefxW76+6SeEhYrcCAF/+I04ntljRGGpa5LCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-06-05T13:36:29.183870Z","bundle_sha256":"c8e170839ecacca0ce86cfeb956a88e9a0afe41535475df5324cb0eef9f20994"}}