Monitoring latent reasoning in looped language models
cs.AI, cs.CL
Submitted: 2026-10-13
Updated: 2026-10-13
Comments: 39 pages, 25 figures, 11 tables
License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
Terminology
Sources
- PonderNet: Learning to Ponder
- Eliciting Latent Predictions from Transformers with the Tuned Lens
- A Mechanistic Analysis of Looped Reasoning Language Models
- Investigating Single-Block Recurrence in Vision Transformers for Image Recognition
- Reasoning Beyond Language: A Comprehensive Survey on Latent Chain-of-Thought Reasoning
- Training Verifiers to Solve Math Word Problems
- Universal Transformers
- Think-at-Hard: Dynamic Looped Transformers for Improved Reasoning
- Adaptive Computation Time for Recurrent Neural Networks
- DeepSeek-R1: Incentivizing Reasoning Capability in LLMs via Reinforcement Learning
- Training Large Language Models to Reason in a Continuous Latent Space
- Measuring Massive Multitask Language Understanding
- Measuring Mathematical Problem Solving With the MATH Dataset
- Chain of Thought Monitorability: A New and Fragile Opportunity for AI Safety
- Latent Chain-of-Thought? Decoding the Depth-Recurrent Transformer
- Prediction Dynamics in Depth-Recurrent Language Models
- HarmBench: A Standardized Evaluation Framework for Automated Red Teaming and Robust Refusal
- Teaching Pretrained Language Models to Think Deeper with Retrofitted Recurrence
- Show Your Work: Scratchpads for Intermediate Computation with Language Models
- Same path, different walk: a mechanistic comparison of looped and stacked transformer encoders on 12-lead ECG
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
- Agents' Last Exam
- 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