Reliable Chain-of-Thought via Prefix Consistency
stat.ML, cs.CL, cs.LG
Submitted: 2026-05-08
Updated: 2026-11-07
Code: https://github.com/naoto-iwase/prefix-consistency
Project page: https://naoto-iwase.github.io/prefix-consistency-page
Terminology
Sources
- Reasoning Theater: Disentangling Model Beliefs from Chain-of-Thought
- Sample, Don't Search: Rethinking Test-Time Alignment for Language Models
- Beyond the Last Answer: Your Reasoning Trace Uncovers More than You Think
- Robust Answers, Fragile Logic: Probing the Decoupling Hypothesis in LLM Reasoning
- Language Models (Mostly) Know What They Know
- Measuring Faithfulness in Chain-of-Thought Reasoning
- Ministral 3
- Nemotron 3 Nano: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
- NVIDIA Nemotron Nano 2: An Accurate and Efficient Hybrid Mamba-Transformer Reasoning Model
- gpt-oss-120b & gpt-oss-20b Model Card
- Inference-Time Computations for LLM Reasoning and Planning: A Benchmark and Insights
- EAGer: Entropy-Aware GEneRation for Adaptive Inference-Time Scaling
- The Sequential Edge: Inverse-Entropy Voting Beats Parallel Self-Consistency at Matched Compute
- Knowledge-Driven CoT: Exploring Faithful Reasoning in LLMs for Knowledge-intensive Question Answering
- FrontierScience: Evaluating AI's Ability to Perform Expert-Level Scientific Tasks
- Benchmarking Reasoning Robustness in Large Language Models
- Path-Consistency with Prefix Enhancement for Efficient Inference in LLMs
Related papers
- Uniform-in-Time Weak Propagation-of-Chaos in Shallow Neural Networks
- Diffusion Models Observe Only Gradients: A Geometric Perspective on Score Matching Errors
- Quantifying Error Propagation and Model Collapse in Diffusion Models
- Behavior of prediction performance metrics with rare events
- Statistical Inference and Learning for Shapley Additive Explanations (SHAP)
- Control Variate Design for Score-Function Gradient Estimators in Importance-Weighted Variational Inference