Reproduce “Fine-tuning LLMs for Tourist Trajectory Prediction using Field Experiment Data” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Tatsuya Amano, Hirozumi Yamaguchi
The authors fine-tune Llama-3.1-8B on visitor trajectories and local context to predict the next point of interest at Wakayama Castle Park. On a held-out test of 113 tourists from 566 trajectories, it achieved 49.1% Accuracy@1, compared with 39.8% for Llama-3-Swallow under the same fine-tuning settings. This measures prediction of observed trips, not effects of hypothetical mobility interventions.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.