The Circuit Basis of Metacognitive Failure in Language Models: Interference and Hallucination
cs.AI
Submitted: 2026-05-07
Updated: 2026-10-31
Terminology
Sources
- The Bayesian Geometry of Transformer Attention
- Is Random Attention Sufficient for Sequence Modeling? Disentangling Trainable Components in the Transformer
- A mathematical perspective on Transformers
- Training Compute-Optimal Large Language Models
- Scaling Laws for Neural Language Models
- Semantic Entropy Probes: Robust and Cheap Hallucination Detection in LLMs
- Universal One-third Time Scaling in Learning Peaked Distributions
- The Spike, the Sparse and the Sink: Anatomy of Massive Activations and Attention Sinks
- Siren's Song in the AI Ocean: A Survey on Hallucination in Large Language Models
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