RecPFN: Prior-Fitted Networks for In-Context-Based Recommendations

Abstract

We introduce RecPFN, a prior-fitted network that brings in-context learning to sequential recommendation. RecPFN is pretrained entirely on synthetic clickstream environments sampled from a broad structural causal prior, enabling it to amortize Bayesian-style inference from a small support set. At inference, a lightweight decoder-only transformer conditions on a handful of domain sequences and produces next-item predictions for queries in a single forward pass, without any weight updates. Across eight public benchmarks, RecPFN achives state-of-the-art zero-shot performance while remaining strongly competitive with supervised methods in low-compute and low-data regimes. It is deployment-efficient and robust to domain shift, outperforming strong zero-shot baselines that rely on large real-interaction corpora. RecPFN provides a practical path toward generalizable, data-efficient recommenders and opens avenues for richer priors, longer-context ICL, and multimodal extensions. Code for training and evaluation is publicly available at https://github.com/SAP-samples/tabular-ai-recpfn/.

12 pages, 4 figures, 8 tables

References in corpus (9)

• S^3-Rec: Self-Supervised Learning for Sequential Recommendation with Mutual Information Maximization
• A Survey of Graph Neural Networks for Social Recommender Systems
• CauseRec: Counterfactual User Sequence Synthesis for Sequential Recommendation
• MMRec: Simplifying Multimodal Recommendation
• On the Effectiveness of Sampled Softmax Loss for Item Recommendation
• Dataset Regeneration for Sequential Recommendation
• A Pre-trained Sequential Recommendation Framework: Popularity Dynamics for Zero-shot Transfer
• LLM-RecG: A Semantic Bias-Aware Framework for Zero-Shot Sequential Recommendation
• Synthetic sequence generator for recommender systems - memory biased random walk on sequence multilayer network