{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/evaluating-rewards-for-question-generation","title":"Evaluating Rewards for Question Generation Models","arxiv_id":"1902.11049","date":"2019-02-28","proceeding":"NAACL 2019 6","authors":["Tom Hosking","Sebastian Riedel"],"abstract":"Recent approaches to question generation have used modifications to a Seq2Seq architecture inspired by advances in machine translation. Models are trained using teacher forcing to optimise only the one-step-ahead prediction. However, at test time, the model is asked to generate a whole sequence, causing errors to propagate through the generation process (exposure bias). A number of authors have proposed countering this bias by optimising for a reward that is less tightly coupled to the training data, using reinforcement learning. We optimise directly for quality metrics, including a novel approach using a discriminator learned directly from the training data. We confirm that policy gradient methods can be used to decouple training from the ground truth, leading to increases in the metrics used as rewards. We perform a human evaluation, and show that although these metrics have previously been assumed to be good proxies for question quality, they are poorly aligned with human judgement and the model simply learns to exploit the weaknesses of the reward source.","url_abs":"https://arxiv.org/abs/1902.11049v2","url_pdf":"https://arxiv.org/pdf/1902.11049v2.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"evaluating-rewards-for-question-generation","repo_url":"https://github.com/bloomsburyai/question-generation","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"machine-translation","task_name":"Machine Translation"},{"task_slug":"policy-gradient-methods","task_name":"Policy Gradient Methods"},{"task_slug":"question-generation","task_name":"Question Generation"},{"task_slug":"question-generation","task_name":"Question-Generation"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"translation","task_name":"Translation"}],"methods":[{"method_slug":"lstm","method_name":"LSTM"},{"method_slug":"seq2seq","method_name":"Seq2Seq"},{"method_slug":"sigmoid-activation","method_name":"Sigmoid Activation"},{"method_slug":"tanh-activation","method_name":"Tanh Activation"}],"datasets_introduced":[],"methods_introduced":[],"results":[{"leaderboard":"/sota/question-generation-on-squad11","task":"Question Generation","dataset":"SQuAD1.1","model":"RNN +attn +copy","rank_in_archive_order":12,"of":13,"metrics":{"BLEU-4":"13.5"},"uses_additional_data":false}],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1902.11049","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}