{"id":"clm_CBD853779110","entity":"openai","predicate":"released","value":"process supervision","value_json":{"feed":"openai-news","item":"https://openai.com/index/improving-mathematical-reasoning-with-process-supervision","tags":{"v":2,"ai":true,"by":"claude-haiku-5-5","on":"2026-11-03","themes":["reasoning"],"release_kind":"research"},"check":{"on":"2026-11-04","url":"https://openai.com/index/improving-mathematical-reasoning-with-process-supervision","read":"feed","pages":1,"quote":"We’ve trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome supervision”).","outcome":"checked"},"evidence":"We’ve trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome supervision”). In addition to boosting performance relative to outcome supervisio","extracted_by":"claude-haiku-5-5","primary_source":true},"valid_from":"2023-05-31","valid_to":null,"published":"2023-05-31","recorded":"2026-11-03","confidence":0.7,"status":"valid","assertable":true,"superseded_by":null,"supersedes":[],"first_seen_at":"2026-11-03T23:55:38.749992+00:00","updated_at":"2026-11-04T01:09:03.345166+00:00","sources":[{"url":"https://openai.com/index/improving-mathematical-reasoning-with-process-supervision","tier":"primary","publisher":"OpenAI","published_at":"2023-05-31T00:00:00+00:00"}],"evidence":{"by":"page-read","on":"2026-11-04","quote":"We’ve trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome supervision”).","url":"https://openai.com/index/improving-mathematical-reasoning-with-process-supervision","read":"feed","excerpt":"We’ve trained a model to achieve a new state-of-the-art in mathematical problem solving by rewarding each correct step of reasoning (“process supervision”) instead of simply rewarding the correct final answer (“outcome supervision”). In addition to boosting performance relative to outcome supervisio"},"legend":{"dates":"valid_from: when the claim took effect (a period figure: the day its period closed). published: the earliest day a source on the record published it (a filing's filing date, a post's date; null when no source carries one). recorded: the day AsOf first held it. A figure can take effect weeks before it is published.","why_not":{},"checked_by":{"page-read":"its source page states the value and the date"}}}