Hi, I've been following this project with great interest and noticed something curious: the scores of AI-generated articles seem to be increasing over time. I'm wondering — is this intentional? Are you continuously retraining the model, or are the human review results being fed back as training signals to improve it? If so, does this mean the AI is actively learning from the reviewers' judgments? Understanding the feedback loop here would help me better interpret the scoring trends. Thanks for the transparency and for running such a fascinating experiment!
Contact: molly030628@gmail.com
Hi, I've been following this project with great interest and noticed something curious: the scores of AI-generated articles seem to be increasing over time. I'm wondering — is this intentional? Are you continuously retraining the model, or are the human review results being fed back as training signals to improve it? If so, does this mean the AI is actively learning from the reviewers' judgments? Understanding the feedback loop here would help me better interpret the scoring trends. Thanks for the transparency and for running such a fascinating experiment!
Contact: molly030628@gmail.com