paper-with-me

홈 › Papers

Epistemic Integrity in Large Language Models

2024-11-10 · Bijean Ghafouri, Shahrad Mohammadzadeh, James Zhou, Pratheeksha Nair, Jacob-Junqi Tian, Hikaru Tsujimura, Mayank Goel, Sukanya Krishna, Reihaneh Rabbany, Jean-François Godbout, Kellin Pelrine

Large language models are increasingly relied upon as sources of information, but their propensity for generating false or misleading statements with high confidence poses risks for users and society. In this paper, we confront the critical problem of epistemic miscalibration $\unicode{x2013}$ where a model's linguistic assertiveness fails to reflect its true internal certainty. We introduce a new human-labeled dataset and a novel method for measuring the linguistic assertiveness of Large Language Models (LLMs) which cuts error rates by over 50% relative to previous benchmarks. Validated across multiple datasets, our method reveals a stark misalignment between how confidently models linguistically present information and their actual accuracy. Further human evaluations confirm the severity of this miscalibration. This evidence underscores the urgent risk of the overstated certainty LLMs hold which may mislead users on a massive scale. Our framework provides a crucial step forward in diagnosing this miscalibration, offering a path towards correcting it and more trustworthy AI across domains.

📄 PDF Abstract BibTeX arXiv:2411.06528

Code (1)

complexdata-mila/epistemic-integrity 공식 구현 pytorch

Similar Papers 제목 키워드 기반

When Helpfulness Becomes Sycophancy: Sycophancy is a Boundary Failure Between Social Alignment and Epistemic Integrity in Large Language Models

2026-05-06 · Jiechen Li, Catherine A. Barry, Rishika Randev, Janet Chen 외 arxiv

This position paper argues that sycophancy in LLMs is a boundary failure between social alignment and epistemic integrity. Existing work often operationalizes sycophancy through external behavior such as agreement with i…

The Polite Liar: Epistemic Pathology in Language Models

2025-11-08 · Bentley DeVilling arxiv

Large language models exhibit a peculiar epistemic pathology: they speak as if they know, even when they do not. This paper argues that such confident fabrication, what I call the polite liar, is a structural consequence…

Reinforcement Learning

Beyond Prediction -- Structuring Epistemic Integrity in Artificial Reasoning Systems

2025-06-19 · Craig Steven Wright

This paper develops a comprehensive framework for artificial intelligence systems that operate under strict epistemic constraints, moving beyond stochastic language prediction to support structured reasoning, proposition…

Knowledge Graphs

LLM Psychosis: A Theoretical and Diagnostic Framework for Reality-Boundary Failures in Large Language Models

2026-04-07 · Ashutosh Raj arxiv

The deployment of large language models (LLMs) as interactive agents has exposed a category of behavioral failure that prevailing terminology, principally hallucination, fails to adequately characterize. This paper intro…

Semantic Reward Collapse and the Preservation of Epistemic Integrity in Adaptive AI Systems

2026-05-12 · William Parris arxiv

Recent advances in reinforcement learning from human feedback (RLHF) and preference optimization have substantially improved the usability, coherence, and safety of large language models. However, recurring behaviors suc…

Reinforcement Learning