Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judge (2025.emnlp-main)
Copied to clipboard
Tianhao Wu, Weizhe Yuan, Olga Golovneva, Jing Xu, Yuandong Tian, Jiantao Jiao, Jason E Weston, Sainbayar Sukhbaatar
| Challenge: | Existing methods for improving large language models have focused on improving model responses rather than judgment capabilities, resulting in rapid saturation during iterative training. |
| Approach: | They propose an iterative Meta-Rewarding step where the model judges its own judgements and uses that feedback to refine its judgment skills. |
| Outcome: | The proposed model improves Llama-3-8B-Instruct from 22.9% to 39.4% on AlpacaEval 2 and 20.6% to 29.1% on Arena-Hard. |
Similar Papers
Unlocking LLMs’ Self-Improvement Capacity with Autonomous Learning for Domain Adaptation (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing models that use self-supervised and instruction fine-tuning can be trained using unlabeled corpora. |
| Approach: | They propose to use unlabeled target corpora to adapt large language models to new domains . they propose to employ self-supervised pre-training and instruction fine-tuning methods . |
| Outcome: | The proposed model can adapt to new domains using only a large amount of unlabeled target corpora. |
Process-based Self-Rewarding Language Models (2025.findings-acl)
Copied to clipboard
| Challenge: | Existing methods to reward LLMs' outputs are not effective in mathematical reasoning scenarios and may lead to a decline in performance. |
| Approach: | They propose a process-based self-rewarding pipeline that integrates long-thought reasoning, step-wise LLM-as-a-Judge, and step- wise preference optimization within the existing paradigm. |
| Outcome: | The proposed model improves the performance of Large Language Models on multiple mathematical reasoning benchmarks and shows that it can surpass human capabilities. |
Teaching Language Models to Self-Improve by Learning from Language Feedback (2024.findings-acl)
Copied to clipboard
| Challenge: | Recent advances in Large Language Models (LLMs) generate content that can be untruthful or harmful. |
| Approach: | They propose a method that leverages model feedback for alignment . they use a base language model to generate initial responses, critiqued and refined . |
| Outcome: | The proposed method outperforms strong baselines across diverse tasks and model sizes. |
Self-Improvement in Multimodal Large Language Models: A Survey (2025.findings-emnlp)
Copied to clipboard
| Challenge: | Using data and data, self-improvement for Large Language Models has improved model capabilities without significantly increasing costs. |
| Approach: | This survey provides a comprehensive overview of self-improvement for Large Language Models . it includes commonly used evaluations and downstream applications . |
| Outcome: | The authors provide a comprehensive overview of self-improvement in Multimodal LLMs. |
Reasons to Reject? Aligning Language Models with Judgments (2024.findings-acl)
Copied to clipboard
| Challenge: | a new framework for aligning large language models with judgments is proposed to help with alignment . a framework that allows for fine-grained inappropriate content detection and correction based on judgments . large language model alignment is critical for making artificial intelligence a reliable ally for humanity . |
| Approach: | They propose a framework that allows for fine-grained inappropriate content detection and correction based on judgments. |
| Outcome: | The proposed framework beats the 175B DaVinci003 and improves on AlpacaEval using judgments. |
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. |
Aligning Large Language Models via Fully Self-Synthetic Data (2026.acl-long)
Copied to clipboard
| Challenge: | Existing approaches to reinforcement learning from human feedback (RLHF) require expensive human-annotated datasets and proprietary models like GPT-4 to annotate preference pairs. |
| Approach: | They propose a self-synthetic framework for LLM alignment where all training data, including prompts (i.e., user queries), responses, and preferences, are generated by the model itself. |
| Outcome: | The proposed framework enhances the model’s chat capabilities on standard benchmarks like AlpacaEval 2.0 while maintaining strong performance on downstream objective tasks. |
Aligning Black-box Language Models with Human Judgments (2025.findings-naacl)
Copied to clipboard
| Challenge: | Large language models (LLMs) are increasingly used as automated judges to evaluate recommendation systems, search engines, and other subjective tasks. |
| Approach: | They propose a framework to align LLM judgments with individual human evaluators or their aggregated judgments without retraining or fine-tuning the LLM. |
| Outcome: | The proposed framework achieves 142% improvement in agreement across 29 tasks and exceeds inter-human agreement on four out of six tasks. |
Calibrating LLM-Based Evaluator (2024.lrec-main)
Copied to clipboard
Yuxuan Liu, Tianchi Yang, Shaohan Huang, Zihan Zhang, Haizhen Huang, Furu Wei, Weiwei Deng, Feng Sun, Qi Zhang
| Challenge: | Existing models for large language models lack the ability to calibrate their outputs towards human preference. |
| Approach: | They propose a multi-stage, gradient-free approach to calibrate an LLM-based evaluator toward human preference. |
| Outcome: | The proposed approach improves correlation with expert evaluation on multiple text quality evaluation datasets. |
Self-Generated Critiques Boost Reward Modeling for Language Models (2025.naacl-long)
Copied to clipboard
Yue Yu, Zhengxing Chen, Aston Zhang, Liang Tan, Chenguang Zhu, Richard Yuanzhe Pang, Yundi Qian, Xuewei Wang, Suchin Gururangan, Chao Zhang, Melanie Kambadur, Dhruv Mahajan, Rui Hou
| Challenge: | Existing reward models produce scalar scores and struggle to incorporate critiques in a natural language format. |
| Approach: | They propose a framework that predicts critiques and rewards using self-generated critiques without extra supervision. |
| Outcome: | The proposed framework improves reward modeling accuracy by 3.7%-7.3% compared to standard reward models and LLM judges. |