We present {that a} GPT-3 mannequin can study to specific uncertainty about its personal solutions in pure language—with out use of mannequin logits. When given a query, the mannequin generates each a solution and a stage of confidence (e.g. “90% confidence” or “excessive confidence”). These ranges map to chances which can be effectively calibrated. The mannequin additionally stays reasonably calibrated below distribution shift, and is delicate to uncertainty in its personal solutions, fairly than imitating human examples. To our data, that is the primary time a mannequin has been proven to specific calibrated uncertainty about its personal solutions in pure language. For testing calibration, we introduce the CalibratedMath suite of duties. We examine the calibration of uncertainty expressed in phrases (“verbalized likelihood”) to uncertainty extracted from mannequin logits. Each sorts of uncertainty are able to generalizing calibration below distribution shift. We additionally present proof that GPT-3’s potential to generalize calibration relies on pre-trained latent representations that correlate with epistemic uncertainty over its solutions.
