[KongchangAI]
BenchmarkDISTRACT基准

DISTRACT

A benchmark for quantifying view bias in learned expression encoders. It turns the previously overlooked problem of view bias into a quantifiable and comparable set of metrics.

Timeline (last 90 days)

Oct 19

The authors built the DISTRACT benchmark to quantify the view-bias phenomenon.

Unverified50%

All Facts (1)

Source Articles