{
  "schema_version": "0.3",
  "id": "deep-reinforcement-learning-from-human-preferences",
  "type": "work",
  "label": "Deep Reinforcement Learning from Human Preferences",
  "aliases": [],
  "description": "scientific article published in January 2017",
  "tier": 2,
  "status": "active",
  "redirect_to": null,
  "fp_kind": null,
  "synthetic": false,
  "identifiers": {
    "wikidata": "Q44650247"
  },
  "inbound_counts": {},
  "claims": [
    {
      "id": "k_1c086649ab81b6914bf33c7b1d9eec95",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "identifier",
      "object": null,
      "value": {
        "scheme": "wikidata",
        "value": "Q44650247"
      },
      "datatype": "identifier",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": "Q44650247",
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.id",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_d80d5cb748536f5eb0ae82ec2ecd6266",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "instance_of",
      "object": {
        "id": "scholarly-article",
        "label": "scholarly article",
        "type": "concept"
      },
      "value": null,
      "datatype": "entity",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": null,
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P31[0].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_e1e31138a4006f22e770b297e63b99ae",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "author",
      "object": {
        "id": "paul-christiano",
        "label": "Paul Christiano",
        "type": "person"
      },
      "value": null,
      "datatype": "entity",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": "Q64769299",
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P50[0].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_46bfe8511db0f4e486c83c99c4b85ff7",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "author",
      "object": {
        "id": "jan-leike",
        "label": "Jan Leike",
        "type": "person"
      },
      "value": null,
      "datatype": "entity",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": "Q123130693",
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P50[1].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_c050ff6a67d5f63c59e14ee86678a386",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "author",
      "object": {
        "id": "shane-legg",
        "label": "Shane Legg",
        "type": "person"
      },
      "value": null,
      "datatype": "entity",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": "Q22278829",
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P50[4].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_918a2733c20abaf540bc07eb75b8b7af",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "author",
      "object": {
        "id": "dario-amodei",
        "label": "Dario Amodei",
        "type": "person"
      },
      "value": null,
      "datatype": "entity",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": "Q103335665",
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P50[5].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    },
    {
      "id": "k_e49b2e2ea23a93ec7526cec6de3c49f4",
      "revision": 1,
      "subject": "deep-reinforcement-learning-from-human-preferences",
      "predicate": "inception",
      "object": null,
      "value": "2017",
      "datatype": "date",
      "polarity": "positive",
      "epistemic": "asserted",
      "temporal": "unknown",
      "status": "current",
      "confidence": 0.7,
      "confidence_method": "tier_default/v1",
      "valid_from": null,
      "valid_to": null,
      "asserted_at": "2026-10-30T17:18:00Z",
      "retrieved_at": "2026-10-30T17:18:00Z",
      "evidence": [
        {
          "url": "https://www.wikidata.org/wiki/Special:EntityData/Q44650247.json",
          "capture": {
            "store": "ev",
            "sha256": "bf192becf852a6745c80c186157049a0650da238e493b670a83eb8183e4f2b69"
          },
          "retrieved_at": "2026-10-30T17:18:00Z",
          "published_at": null,
          "quote": null,
          "tier": "wiki",
          "locator": "jsonpath:$.entities.Q44650247.claims.P577[0].mainsnak.datavalue.value",
          "media": null,
          "source_family": "wikidata",
          "method": "api",
          "extractor": {
            "name": "build_release",
            "version": "0.3.1"
          },
          "rights": {
            "holder": "Wikidata contributors",
            "license": "CC0-1.0",
            "basis": "license",
            "display": true,
            "export": true,
            "train": true,
            "policy_version": "2026-10-20"
          }
        }
      ],
      "derived_from": [],
      "derivation": null,
      "qualifiers": {},
      "negation": null,
      "supersedes": null,
      "mirror_of": null
    }
  ],
  "media": [],
  "version": 1,
  "created_at": "2026-10-30T17:18:00Z",
  "updated_at": "2026-10-30T17:18:00Z",
  "content_hash": "0231604b42482385ce4ab0b13a1efdef980867e79bceeaba1d831e55aa2c4076",
  "license": "ENTITI-ATTR-1.0"
}