A fine-tuned DistilRoBERTa recipe-quality predictor achieves a test-set MAE of 0.39, slightly better than but not significantly different from a mean-prediction baseline of 0.42.¶

Claim · All claims
Evidence · 1 study · causal r2 · q2
Currency · newest study 2025 · not yet checked against its sources · not reviewed by a person

Subclaims¶

q2 i? RELEXset-Predictor achieved a mean absolute error of 0.39 on the hold-out test set, slightly better than the 0.42 baseline MAE, with the difference not significant. → Mejia-Domenzain 2025

Evidence¶

Mejia-Domenzain 2025¶

Mejia-Domenzain, P., Frej, J., Neshaei, S. P., Mouchel, L., Nazaretsky, T., Wambsganss, T., Bosselut, A., & Käser, T. (2025). Enhancing Procedural Writing Through Personalized Example Retrieval: A Case Study on Cooking Recipes. Int J Artif Intell Educ. https://doi.org/10.1007/s40593-024-00405-1

q2 · i? · causal · r2

Numerical evaluation of the fine-tuned DistilRoBERTa regression model (RELEXset-Predictor) on a held-out test split of the rated-recipe database. As printed, "slightly better than the baseline MAE of 0.42 (simply predicting the mean)", with the article noting the difference was not significant. No effect size is printed.

"RELEXset-Predictor achieved a mean absolute error (MAE) of 0 .39 on the test set, which is slightly better than the baseline MAE of 0.42 (simply predicting the mean)."

Discussion¶

Related Claims¶

• Under identical LoRA hyperparameters and training data, the 27B Gemma model outperforms the 70B LLaMA model on every computed essay-scoring metric — related
• Jointly predicting courses and grades reduces grade-prediction MAE by 46.4% versus an identical Transformer that predicts grades only — related