TableBench: A Comprehensive and Complex Benchmark for Table Question Answering

Open paper

Benchmark	Model	Rank	Results
table-question-answering-on-tablebench	GPT-4-Turbo (TCoT)	#1	Score: 51.32
table-question-answering-on-tablebench	GPT-4o (TCoT)	#2	Score: 50.39
table-question-answering-on-tablebench	Llama3.1-70B (TCoT)	#3	Score: 43.85
table-question-answering-on-tablebench	TableLLM-Llama3.1-8B (TCoT)	#4	Score: 35.29
table-question-answering-on-tablebench	GPT-3.5-Turbo (TCoT)	#5	Score: 30.85