The Alignment Paradox: How Post-Training Amplifies Confident Hallucinations in Language Models
cs.CL, cs.AI
Submitted: 2026-09-26
Updated: 2026-09-26
Code: https://github.com/star5o/HCE
Terminology
Sources
- Concrete Problems in AI Safety
- Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback
- Eliciting Latent Predictions from Transformers with the Tuned Lens
- Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge
- Training Verifiers to Solve Math Word Problems
- The Llama 3 Herd of Models
- KTO: Model Alignment as Prospect Theoretic Optimization
- Mistral 7B
- Language Models (Mostly) Know What They Know
- DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models
- Llama 2: Open Foundation and Fine-Tuned Chat Models
- Qwen2 Technical Report
Related papers
- Exploring Solution Divergence and Its Effect on Large Language Model Problem Solving
- Ishigaki-IDS-Bench: A Benchmark for Generating Information Delivery Specification from BIM Information Requirements
- Subliminal Steering: Stronger Encoding of Hidden Signals
- MedStruct-S: A Benchmark for Key Discovery, Key-Conditioned QA and Semi-Structured Extraction from OCR Clinical Reports
- The End of Transformers? On Challenging Attention and the Rise of Sub-Quadratic Architectures
- Untangling the Mechanisms of Misleading Context in Medical Question Answering