ResearchBenchmarksTabular Classification on Tabular Classification Benchmarks 256-shotFollow76.6Average Test AccuracyBoostLLM56.73661.89367.0572.207May 7, 2026Evaluation ResultsMethodMethodLinksAverage Test AccuracyBoostLLMLLM Used=Qwen3-4BLLM Used=Qwen3-4B2026.0576.6DeLTaLLM Used=GPT-4oLLM Used=GPT-4o2026.0570TP-BERTaLLM Used=RoBERTaLLM Used=RoBERTa2026.0564.9FeatLLMLLM Used=GPT-4oLLM Used=GPT-4o2026.0563.9LIFTLLM Used=GPT-3.5LLM Used=GPT-3.52026.0561TabLLMLLM Used=T0LLM Used=T02026.0557.5