Challenge: Recent studies have shown that large language models can solve complex reasoning tasks with Chain-of-Thought Prompting.
Approach: They propose a training method where the LLM is tasked to explore the first wrong step within the rationale and use such signals as fine-grained rewards for further improvement.
Outcome: The proposed model improves on the GSM8K and MATH test sets by 11.57% and 2.89% on average compared to supervised fine-tuning (SFT).

Similar Papers

Large Language Models Can Self-Improve (2023.emnlp-main)

Copied to clipboard

Challenge: Large Language Models (LLMs) have excellent performance in various tasks, but fine-tuning requires extensive supervision.
Approach: They propose to use a pre-trained Large Language Model to generate rationale-augmented answers for unlabeled questions and fine-tune the LLM using those self-generated solutions as target outputs.
Outcome: The proposed approach improves the general reasoning ability of a 540B-parameter LLM without any ground truth label.
Self-Training Elicits Concise Reasoning in Large Language Models (2025.findings-acl)

Copied to clipboard

Challenge: Chain-of-thought reasoning has enabled large language models to use additional computation through intermediate tokens to solve complex tasks, but current models often generate more tokens than necessary to accomplish the task, incurring extraneous inference costs.
Approach: They propose to fine-tune models with self-generated concise reasoning paths obtained by best-of-N sampling and few-shot conditioning in task-specific settings to elicit concise reasoning.
Outcome: The proposed method reduces output tokens by 30% on GSM8K and MATH while maintaining average accuracy.
ReFT: Reasoning with Reinforced Fine-Tuning (2024.acl-long)

Copied to clipboard

Challenge: Existing approaches to improve the generalization of large language models are using Supervised Fine-Tuning (SFT) this approach does not show sufficient generalization ability because it only relies on the given CoT data.
Approach: They propose to use Chain-of-Thought annotations to train Large Language Models using supervised fine-tuning to improve generalization.
Outcome: The proposed approach outperforms SFT on GSM8K, MathQA, and SVAMP datasets and shows a superior generalization ability.
Unleashing the Reasoning Potential of LLMs by Critique Fine-Tuning on One Problem (2025.emnlp-main)

Copied to clipboard

Challenge: Critique Fine-Tuning (CFT) is a promising paradigm for unlocking the reasoning capabilities of large language models.
Approach: They propose a method that leverages critique data generated from a single math problem to improve reasoning accuracy.
Outcome: The proposed method surpasses one-shot RLVR while requiring 15 to 20 times less compute.
Can Large Language Models Invent Algorithms to Improve Themselves? (2025.naacl-long)

Copied to clipboard

Challenge: Large Language Models (LLMs) have shown remarkable performance improvements, but the methods for improving LLMs are still designed by humans.
Approach: They propose a framework which enables LLMs to generate and learn model-improvement algorithms by the seed model.
Outcome: The proposed framework outperforms human-designed methods in model-improving tasks and improves the seed model by 6% and outperformed human-design methods by 4.3% on GSM8k.
Large Language Models are Better Reasoners with Self-Verification (2023.findings-emnlp)

Copied to clipboard

Challenge: Existing methods to solve complex natural language processing tasks require multiple steps to verify the answers.
Approach: They propose to use chain of thought prompting to solve reasoning tasks with large language models.
Outcome: The proposed method can improve reasoning performance on arithmetic, commonsense, and logical reasoning datasets.
ChatGLM-Math: Improving Math Problem-Solving in Large Language Models with a Self-Critique Pipeline (2024.findings-emnlp)

Copied to clipboard

Challenge: Large language models (LLMs) have shown excellent mastering of human language but struggle in real-world applications that require mathematical problem-solving.
Approach: They propose a pipeline to train a general Math-Critique model from the LLM itself to provide feedback signals and employ rejective fine-tuning and direct preference optimization over the Llm's own generations for data collection.
Outcome: The proposed pipeline outperforms existing LLMs that could be two times larger.
MinT: Boosting Generalization in Mathematical Reasoning via Multi-view Fine-tuning (2024.lrec-main)

Copied to clipboard

Challenge: Existing methods focus on specializing LMs in mathematical reasoning and rely on knowledge distillation.
Approach: They propose a multi-view fine-tuning method that exploits existing mathematical problem datasets with diverse annotation styles.
Outcome: The proposed method outperforms existing methods that rely heavily on LLM teachers . it grants models generalization ability across views and datasets, and the capability to learn from inaccurate or incomplete data.
Improving Language Model Reasoning with Self-motivated Learning (2024.lrec-main)

Copied to clipboard

Challenge: Large-scale high-quality training data is important for improving the performance of models.
Approach: They propose a framework that motivates the model to automatically generate rationales on existing datasets and improves the performance of reasoning through reinforcement learning.
Outcome: The proposed model outperforms InstructGPT on multiple reasoning datasets and outperformed InstructGPT on other datasets.
Mitigating Tail Narrowing in LLM Self-Improvement via Socratic-Guided Sampling (2025.naacl-long)

Copied to clipboard

Challenge: Large language models (LLMs) generate solutions themselves and iteratively train on filtered, high-quality rationales, but performance reaches a ceiling after a few iterations.
Approach: They propose a strategy to improve the efficiency of sampling heavy-tailed data by using Socratic-style guidance signals to help LLMs reasoning with complex queries.
Outcome: The proposed approach is effective on difficult queries and on held-out tasks, while requiring human supervision.

What is GenGO?

GenGO is an NLP powered publication search system. It currenctly indexes 30k+ papers from ACL Anthology, and implements multi-aspect summarization, semantic search, and more!

Information

About
Limitations