StepBack: Step-Level Error-Localized Resampling for Efficient Test-Time Reasoning
cs.AI, cs.CL, cs.LG
Submitted: 2026-11-05
Updated: 2026-11-05
Code: https://github.com/JEKimLab/StepBack
Terminology
Sources
- Language Models are Few-Shot Learners
- Training Verifiers to Solve Math Word Problems
- Skywork Open Reasoner 1 Technical Report
- Measuring Mathematical Problem Solving With the MATH Dataset
- The First Few Tokens Are All You Need: An Efficient and Effective Unsupervised Prefix Fine-Tuning Method for Reasoning Models
- Large Language Models are Zero-Shot Reasoners
- Deductive Verification of Chain-of-Thought Reasoning
- Adaptive Rectification Sampling for Test-Time Compute Scaling
- Solving math word problems with process- and outcome-based feedback
- Self-Consistency Improves Chain of Thought Reasoning in Language Models
- Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Learning From Correctness Without Prompting Makes LLM Efficient Reasoner
- The Lessons of Developing Process Reward Models in Mathematical Reasoning
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