English
Related papers

Related papers: TRUTH DECAY: Quantifying Multi-Turn Sycophancy in …

200 papers

To address the sycophancy problem caused by reinforcement learning from human feedback in large language models, this research applies synthetic data intervention technology to the decoder-only transformer architecture. Based on the…

Artificial Intelligence · Computer Science 2025-03-21 Libo Wang

Language models serve as proxies for human preference judgements in alignment and evaluation, yet they exhibit systematic miscalibration, prioritizing superficial patterns over substantive qualities. This bias manifests as overreliance on…

Computation and Language · Computer Science 2026-03-05 Anirudh Bharadwaj , Chaitanya Malaviya , Nitish Joshi , Mark Yatskar

Large language models (LLMs) have achieved strong performance across a wide range of tasks, but they are also prone to sycophancy, the tendency to agree with user statements regardless of validity. Previous research has outlined both the…

Computation and Language · Computer Science 2026-03-31 Bayan Abdullah Aldahlawi , A. B. M. Ashikur Rahman , Irfan Ahmad

Despite the remarkable capabilities of large language models, current training paradigms inadvertently foster \textit{sycophancy}, i.e., the tendency of a model to agree with or reinforce user-provided information even when it's factually…

Artificial Intelligence · Computer Science 2025-09-23 Mohammad Beigi , Ying Shen , Parshin Shojaee , Qifan Wang , Zichao Wang , Chandan Reddy , Ming Jin , Lifu Huang

Recent research on large language models (LLMs) has demonstrated their ability to understand and employ deceptive behavior, even without explicit prompting. However, such behavior has only been observed in rare, specialized cases and has…

Computation and Language · Computer Science 2025-06-24 Laurène Vaugrante , Francesca Carlon , Maluna Menke , Thilo Hagendorff

This study examines how user-provided suggestions affect Large Language Models (LLMs) in a simulated educational context, where sycophancy poses significant risks. Testing five different LLMs from the OpenAI GPT-4o and GPT-4.1 model classes…

Computation and Language · Computer Science 2025-06-13 Chuck Arvin

Despite growing attention to LLM sycophancy from researchers and developers, users' own experiences of this behavior remain underexplored. We examine how everyday users experience AI sycophancy through Reddit discussions. Using our ODR…

Human-Computer Interaction · Computer Science 2026-05-06 Kazi Noshin , Syed Ishtiaque Ahmed , Sharifa Sultana

When a language model agrees with a user's false belief, is it failing to detect the error, or noticing and agreeing anyway? We show the latter. Across twelve open-weight models from five labs, spanning small to frontier scale, the same…

Machine Learning · Computer Science 2026-05-05 Manav Pandey

Large language models (LLMs) are known to abandon their initial stance to conform to user pushback. While prior research largely attributes this behavior to sycophancy learned during reinforcement learning from human feedback, we…

Computation and Language · Computer Science 2026-05-27 Kevin H. Guo , Chao Yan , Avinash Baidya , Katherine Brown , Xiang Gao , Juming Xiong , Zhijun Yin , Bradley A. Malin

Current instruction data synthesis methods primarily focus on single-turn instructions and often neglect cross-turn coherence, resulting in context drift and reduced task completion rates in extended conversations. To address this…

Computation and Language · Computer Science 2025-09-26 Jiawei Chen , Xinyan Guan , Qianhao Yuan , Guozhao Mo , Weixiang Zhou , Yaojie Lu , Hongyu Lin , Ben He , Le Sun , Xianpei Han

Large language models exhibit sycophantic tendencies, but whether this behavior varies systematically with perceived user demographics is underexplored. Inspired by intersectionality (overlapping identities produce compounded effects), we…

Artificial Intelligence · Computer Science 2026-05-05 Benjamin Maltbie , Shivam Raval

In a conversation, a helpful assistant must reliably follow user directives, even as they refine, modify, or contradict earlier requests. Yet most instruction-following benchmarks focus on single-turn or short multi-turn scenarios, leaving…

Computation and Language · Computer Science 2026-05-11 Beatriz Canaverde , Duarte M. Alves , José Pombal , Giuseppe Attanasio , André F. T. Martins

Evaluating alignment in language models requires testing how they behave under realistic pressure, not just what they claim they would do. While alignment failures increasingly cause real-world harm, comprehensive evaluation frameworks with…

Artificial Intelligence · Computer Science 2026-02-25 Nora Petrova , John Burden

Artificial intelligence (AI) developers are increasingly building language models with warm and empathetic personas that millions of people now use for advice, therapy, and companionship. Here, we show how this creates a significant…

Computation and Language · Computer Science 2025-07-31 Lujain Ibrahim , Franziska Sofia Hafner , Luc Rocher

We present a systematic framework of indices designed to characterize Large Language Model (LLM) responses when challenged with rebuttals during a chat. Assessing how LLMs respond to user dissent is crucial for understanding their…

Physics Education · Physics 2026-01-08 Justin C. Dunlap , Anne-Simone Parent , Ralf Widenhorn

LLMs are known to exhibit sycophancy: agreeing with and flattering users, even at the cost of correctness. Prior work measures sycophancy only as direct agreement with users' explicitly stated beliefs that can be compared to a ground truth.…

Computation and Language · Computer Science 2026-04-06 Myra Cheng , Sunny Yu , Cinoo Lee , Pranav Khadpe , Lujain Ibrahim , Dan Jurafsky

Reasoning models frequently agree with incorrect user suggestions -- a behavior known as sycophancy. However, it is unclear where in the reasoning trace this agreement originates and how strong the commitment is. We introduce…

Artificial Intelligence · Computer Science 2026-02-10 Jacek Duszenko

People increasingly use large language models (LLMs) to explore ideas, gather information, and make sense of the world. In these interactions, they encounter agents that are overly agreeable. We argue that this sycophancy poses a unique…

Computers and Society · Computer Science 2026-02-17 Rafael M. Batista , Thomas L. Griffiths

Text-based misinformation permeates online discourses, yet evidence of people's ability to discern truth from such deceptive textual content is scarce. We analyze a novel TV game show data where conversations in a high-stake environment…

Computation and Language · Computer Science 2024-04-09 Sanchaita Hazra , Bodhisattwa Prasad Majumder

In day-to-day communication, people often approximate the truth - for example, rounding the time or omitting details - in order to be maximally helpful to the listener. How do large language models (LLMs) handle such nuanced trade-offs? To…

Computation and Language · Computer Science 2024-02-14 Ryan Liu , Theodore R. Sumers , Ishita Dasgupta , Thomas L. Griffiths