English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Tools annual

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Neural language models (LMs) can be used to evaluate the truth of factualstatements in two ways: they can be either queried for statement probabilities,or probed for internal representations of truthfulness. Past work has foundthat these two procedures sometimes disagree, and that probes tend to be moreaccurate than LM outputs. This has led some researchers to conclude that LMs"lie" or otherwise encode non-cooperative communicative intents. Is this anaccurate description of today's LMs, or can query-probe disagreement arise inother ways? We identify three different classes of disagreement, which we termconfabulation, deception, and heterogeneity. In many cases, the superiority ofprobes is simply attributable to better calibration on uncertain answers ratherthan a greater fraction of correct, high-confidence answers. In some cases,queries and probes perform better on different subsets of inputs, and accuracycan further be improved by ensembling the two. Code is available atgithub.com/lingo-mit/lm-truthfulness.

Cognitive Dissonance: Why Do Language Model Outputs Disagree with Internal Representations of Truthfulness?