Evaluating AI Alignment in LLMs: Output Analysis of Value Priorities Across 75 Models with Human Benchmarking
cs.AI, cs.HC
Submitted: 2025-06-14
Updated: 2026-05-16
Terminology
Sources
- Moral Foundations of Large Language Models
- Foundational Challenges in Assuring Alignment and Safety of Large Language Models
- Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
- Constitutional AI: Harmlessness from AI Feedback
- SaGE: Evaluating Moral Consistency in Large Language Models
- Measuring Progress on Scalable Oversight for Large Language Models
- Open Problems and Fundamental Limitations of Reinforcement Learning from Human Feedback
- Beyond prompt brittleness: Evaluating the reliability and consistency of political worldviews in LLMs
- Sycophancy to Subterfuge: Investigating Reward-Tampering in Large Language Models
- Towards Measuring the Representation of Subjective Global Opinions in Language Models
- AI Control: Improving Safety Despite Intentional Subversion
- Alignment Revisited: Are Large Language Models Consistent in Stated and Revealed Preferences?
- Machine Psychology
- Aligning AI With Shared Human Values
- Learning Fine-Grained Grounded Citations for Attributed Large Language Models
- AI Alignment: A Comprehensive Survey
- Language Models (Mostly) Know What They Know
- What are human values, and how do we align AI to them?
- Legal Alignment for Safe and Ethical AI
- Understanding Critical Thinking in Generative Artificial Intelligence Use: Development, Validation, and Correlates of the Critical Thinking in AI Use Scale
Related papers
- Multi-Objective Structured Pruning of LLMs for Latency and Model Size Optimization
- MAVEN-T: Reinforced Heterogeneous Distillation for Real-Time Multi-Agent Trajectory Prediction
- Model Discovery Agent: LLM-assisted Bayesian experiment design for data-efficient discovery of mechanistic world models
- Embeddings for Preferences, Not Semantics
- Can We Trust Item Response Theory for AI Evaluation?
- Diagnosing Task Insensitivity in Language Agents