{"id":"2605.30251","title":"Same Evidence, Different Answers: Canonical-Context On-Policy Distillation for Multi-Turn Language Models","url":"https://codeofpaper.com/paper/2605.30251","arxiv_url":"https://arxiv.org/abs/2605.30251","published":"2026-05-28","code":"none","best":null,"most_starred_community":null,"implementations":[],"possible_matches":0}