The Commit-Abstain Circuit: Why Language Models Hallucinate Instead of Abstaining
cs.AI
Submitted: 2026-09-26
Updated: 2026-09-26
Code: https://github.com/vnht/commit-abstain-circuit
Terminology
Sources
- Mitigating LLM Hallucinations via Conformal Abstention
- Twin Worlds: Equivariance-Based Abstention for Evidence-Grounded Reasoning
- TruthRL: Incentivizing Truthful LLMs via Reinforcement Learning
- In-context Learning and Induction Heads
- The Atlas of In-Context Learning: How Attention Heads Shape In-Context Retrieval Augmentation
- First Hallucination Tokens Are Different from Conditional Ones
- Qwen3 Technical Report
- The Llama 3 Herd of Models
- Ministral 3
Related papers
- 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
- The Clinician's Veto: Navigating Trust, Liability, and Uncertainty in Autonomous AI Prescribing
- MindHelper: Closed-Loop Embodied Mental-State Reasoning for Precision Intervention
- Incumbent Advantage: Brand Bias and Cognitive Manipulation Dynamics in LLM Recommendation Systems
- VSAL: A Vision Solver with Adaptive Layouts for Graph Property Detection