1 paper · 1 filter
Ahrii Kim, Chanjun Park, Seong-heum Kim
Recent calls for harder machine translation benchmarks have not clarified what difficulty should mean. We argue that one meaningful and currently unmeasured axis is referential rea…