We knowthat many limitations remain as discussed above and we plan to make regular model updates to improve in such areas. Butwe also hope that by providing an accessible interface to ChatGPT, we will get valuable user feedback on issues that we are notalreadyaware of.
Users are encouraged to provide feedback on problematic modeloutputs through the UI, as well as on false positives/negatives fromthe external content filter which is also part of the interface. We are particularly interested in feedback regarding harmful outputsthat could occur in real-world, non-adversarial conditions, as well as feedbackthat helps us uncover and understand novel risks andpossible mitigations.You can choose to enter the ChatGPTFeedback Contest³ for achance to win up to $500 in APIcredits. Entriescan be submitted via the feedback form that is linked in the ChatGPT interface.
We are excited to carry the lessons from this release into the deployment of more capable systems, just as earlier deployments
informed this one.
本文来自知之小站
PDF报告已分享至知识星球，微信扫码加入查阅下载3万+精选资料，年享1万+精选更新
（星球内含更多未发布精选报告.其它事宜可联系zzxz_88@163.com）
