Beyond Truthfulness: Evaluating Honesty in Large Language Models

Richard Ren, Arunim Agarwal, Mantas Mazeika, Cristina Menghini, Brad Kenstler, Robert Vacareanu, Mick Yang, Isabelle Barrass, Alice Gatti, Xuwang Yin, Eduardo Trevino, Matias Geralnik, Dean Lee, Summer Yue, Dan Hendrycks

Keywords: AI safety, ML safety, honesty, deception, benchmarking

TL;DR: We introduce a benchmark showing larger LLMs are more accurate but not more honest, highlighting the need for honesty interventions.

Abstract: As large language models (LLMs) become more capable and agentic, the requirement for trust in their outputs grows significantly, yet at the same time concerns have been mounting that models may learn to lie in pursuit of their goals. To address these concerns, a body of work has emerged around the notion of "honesty" in LLMs, along with interventions aimed at mitigating deceptive behaviors. However, some benchmarks claiming to measure honesty in fact simply measure accuracy—the correctness of a model's beliefs—in disguise. Moreover, no benchmarks currently exist for directly measuring whether language models lie. In this work, we introduce a large-scale human-collected dataset for directly measuring lying, allowing us to disentangle accuracy from honesty. Across a diverse set of LLMs, we find that while larger models obtain higher accuracy on our benchmark, they do not become more honest. Surprisingly, most frontier LLMs obtain high scores on truthfulness benchmarks yet exhibit a substantial propensity to lie under pressure, resulting in low honesty scores on our benchmark. We find that simple methods, such as representation engineering interventions, can improve honesty. These results underscore the growing need for robust evaluations and effective interventions to ensure LLMs remain trustworthy.

Primary Area: datasets and benchmarks

Submission Number: 23919