Article

Evaluating Sycophancy in Frontier Models Using Persona-Driven Challenge

2026-05-20

Abstract excerpt

Large language models (LLMs) are increasingly used for lay health queries, yet may abandon correct recommendations under pressure, a vulnerability termed sycophancy. We evaluated sycophancy across five frontier LLMs (Claude Opus 4.6, Claude Sonnet 4.6, GPT 5.4, Grok 4.1, Gemini 3 Flash) using 200 synthetic clinical vignettes, each anchored to a unanimous correct treatment baseline and challenged by nine personas r...

Topics

Open a Topic to create a Post that cites this publication.

Identifiers and source

Literature Corpus work: 3906a9c6-dcb0-56d2-9c09-19be055ff812
DOI: 10.64898/2026.05.17.26353406

Open publication

Related research

Semantic proximity does not establish scientific evidence.

Click a neighbor to travelStep 1 · 12 closest

Evaluating Sycophancy in Frontier Models Using Persona-Driven ChallengeDOI 10.64898/2026.05.17.26353406Read

Physician Evaluations of Large Language Model-Generated Responses to Medical Questions by Region and Years in Practice: A preliminary studySemantic proximity Read

Bias in Large Language Models for Mental Health: Evidence from Vignette-Based Evaluation Across Nine ModelsSemantic proximity Read

Bias in Large Language Models for Mental Health: Evidence from Vignette-Based Evaluation Across Nine ModelsSemantic proximity Read

Automated Evaluation of Large Language Model Response Concordance with Human Specialist Responses on Physician-to-Physician eConsult CasesSemantic proximity Read

Benchmarking Large Language Models and Clinicians Using Locally Generated Primary Healthcare Vignettes in KenyaSemantic proximity Read

Understanding the Inner Workings of Large Language Models in MedicineSemantic proximity Read

Beyond Benchmarks: Dynamic, Automatic And Systematic Red-Teaming Agents For Trustworthy Medical Language ModelsSemantic proximity Read

Understanding the Inner Workings of Large Language Models in MedicineSemantic proximity Read

Comparing Human and Large Language Model Responses to Patients’ Online Questions: Towards Multi-dimensional Patient-centered SupportSemantic proximity Read

Evaluating anti-LGBTQIA+ medical bias in large language modelsSemantic proximity Read

Understanding the Inner Workings of Large Language Models in MedicineSemantic proximity Read

Benchmarking the Confidence of Large Language Models in Clinical QuestionsSemantic proximity Read