A large-scale benchmark finds that in-domain fine-tuning works best for foundation model recommenders, while cross-dataset and multi-domain training help in new scenarios.
C.2 Additional Evaluation Metrics In the main text, we report only the AUC metric due to the space constraints
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.IR 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Evaluating Recabilities of Foundation Models: A Multi-Domain, Multi-Dataset Benchmark
A large-scale benchmark finds that in-domain fine-tuning works best for foundation model recommenders, while cross-dataset and multi-domain training help in new scenarios.