Our work “Evaluating Attribution in Dialogue Systems: The BEGIN Benchmark” got accepted at TACL 2022 [Paper] .