← back to paper
arxiv: 2605.06117 · 2 revisions
BoostLLM: Boosting-inspired LLM Fine-tuning for Few-shot Tabular Classification