Pith. sign in

arXiv preprint arXiv:2205.00049 , year=

Prompt Consistency for Zero-Shot Task Generalization , author= · arXiv 2205.00049

2 Pith papers cite this work. Polarity classification is still indexing.

2 Pith papers citing it
read on arXiv browse 2 citing papers

fields

cs.CL 2

years

2023 1 2022 1

verdicts

CONDITIONAL 2

representative citing papers

Discovering Latent Knowledge in Language Models Without Supervision

cs.CL · 2022-12-07 · conditional · novelty 8.0

An unsupervised technique extracts latent yes-no knowledge from language model activations by locating a direction that satisfies logical consistency properties, outperforming zero-shot accuracy by 4% on average across models and datasets.

REPLUG: Retrieval-Augmented Black-Box Language Models

cs.CL · 2023-01-30 · conditional · novelty 6.0

REPLUG improves frozen black-box LMs by prepending LM-supervised retrieved documents, delivering 6.3% better language modeling on GPT-3 and 5.1% better five-shot MMLU on Codex.

citing papers explorer

Showing 2 of 2 citing papers.

  • Discovering Latent Knowledge in Language Models Without Supervision cs.CL · 2022-12-07 · conditional · none · ref 36

    An unsupervised technique extracts latent yes-no knowledge from language model activations by locating a direction that satisfies logical consistency properties, outperforming zero-shot accuracy by 4% on average across models and datasets.

  • REPLUG: Retrieval-Augmented Black-Box Language Models cs.CL · 2023-01-30 · conditional · none · ref 207

    REPLUG improves frozen black-box LMs by prepending LM-supervised retrieved documents, delivering 6.3% better language modeling on GPT-3 and 5.1% better five-shot MMLU on Codex.