CycleAlign: Iterative Distillation from Black-box LLM to White-box Models for Better Human Alignment (2024.findings-acl)
Copied to clipboard
| Challenge: | Existing language models that generate harmful responses are constrained by their inherent capability. |
| Approach: | They propose to align large language models with human preferences from AI feedback. |
| Outcome: | The proposed framework improves the alignment of large language models with human preferences from AI feedback. |
Similar Papers
MetaAlign: Align Large Language Models with Diverse Preferences during Inference Time (2025.findings-naacl)
Copied to clipboard
| Challenge: | Existing methods to align large language models with human preferences often result in a static alignment that cannot account for the diversity of human preferences in practical applications. |
| Approach: | They propose a method to help large language models dynamically align with various explicit or implicit preferences specified at inference time. |
| Outcome: | The proposed method can help LLMs dynamically align with various explicit or implicit preferences specified at the inference stage, validating the feasibility of MetaAlign. |
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. |
AutoAlign: Get Your LLM Aligned with Minimal Annotations (2025.acl-demo)
Copied to clipboard
Xinyu Lu, Dong Xu, Chunkang Zhang, Xinyan Guan, Junxiang Wang, Qingyu Zhang, Pengbo Wang, Yingzhi Mao, Hao Xiang, Xueru Wen, Zichao Li, Yaojie Lu, Hongyu Lin, Le Sun, Xianpei Han
| Challenge: | Automated Alignment (ALM) is a set of algorithms designed to align Large Language Models (LLMs) with human intentions and values while minimizing manual intervention. |
| Approach: | They propose an open-source toolkit that integrates mainstream automated algorithms through a consistent interface and an accessible workflow supporting one-click execution for prompt synthesis and automatic alignment signal construction. |
| Outcome: | The proposed framework enables easy reproduction of existing results through extensive benchmarks and facilitates the development of novel approaches via modular components. |
Aligning Large Language Models through Synthetic Feedback (2023.emnlp-main)
Copied to clipboard
| Challenge: | Currently, alignment learning requires significant human demonstrations and feedback from proprietary LLMs such as ChatGPT. |
| Approach: | They propose a framework that uses synthetic feedback to align large language models to human values without extensive human annotations and proprietary LLMs. |
| Outcome: | The proposed model outperforms open-source models on human-annotated demonstrations in alignment benchmarks. |
Don’t Forget Your Reward Values: Language Model Alignment via Value-based Calibration (2024.emnlp-main)
Copied to clipboard
| Challenge: | Existing methods for generating large language models have been criticized for their complexity and instability. |
| Approach: | They propose a value-based calibration method to better align Large Language Models with human preferences. |
| Outcome: | The proposed method surpasses existing methods on AI assistant and summarization datasets, providing impressive generalizability, robustness, and diversity in different settings. |
GenderAlign: An Alignment Dataset for Mitigating Gender Bias in Large Language Models (2025.acl-long)
Copied to clipboard
Tao Zhang, Ziqian Zeng, YuxiangXiao YuxiangXiao, Huiping Zhuang, Cen Chen, James R. Foulds, Shimei Pan
| Challenge: | Large Language Models (LLMs) generate content that exhibits gender biases, raising ethical concerns. |
| Approach: | They propose to use a dataset to identify gender biases in Large Language Models (LLMs) this dataset is a "chosen" and "rejected" LLM alignment is an effective approach to mitigate gender bias. |
| Outcome: | The proposed dataset shows that it reduces gender bias and improves quality. |
Black-Box Prompt Optimization: Aligning Large Language Models without Model Training (2024.acl-long)
Copied to clipboard
| Challenge: | Large language models are often not well aligned with human intents, which requires additional training. |
| Approach: | They propose to use Black-Box Prompt Optimization (BPO) to perform alignments on large language models that are not well aligned with human intents. |
| Outcome: | The proposed model outperforms existing models and is model-agnostic. |
IterAlign: Iterative Constitutional Alignment of Large Language Models (2024.naacl-long)
Copied to clipboard
| Challenge: | Empirical results show that iterAlign improves truthfulness, helpfulness, harmlessness and honesty, improving the LLM alignment by up to 13.5% in harmlessness. |
| Approach: | They propose a data-driven constitution discovery and self-alignment framework called IterAlign to overcome these drawbacks by leveraging red teaming to uncover weaknesses of an LLM. |
| Outcome: | Empirical results show that iterAlign improves truthfulness, helpfulness, harmlessness and honesty by up to 13.5%. |
AlignDistil: Token-Level Language Model Alignment as Adaptive Policy Distillation (2025.acl-long)
Copied to clipboard
| Challenge: | Existing methods for LLM alignment optimize tokens using a sparse, response-level reward or preference annotation. |
| Approach: | They propose an RLHF-equivalent distillation method for token-level reward optimization that incorporates the reward learned by DPO into the RLHG objective and builds a token-based teacher distribution. |
| Outcome: | The proposed method bridges the accuracy gap between the reward from the DPO model and the pure reward model by building a contrastive DPO reward with a normal and a reverse DPO. |
Aligning Large Language Models with Human Preferences through Representation Engineering (2024.acl-long)
Copied to clipboard
Wenhao Liu, Xiaohua Wang, Muling Wu, Tianlong Li, Changze Lv, Zixuan Ling, Zhu JianHao, Cenyuan Zhang, Xiaoqing Zheng, Xuanjing Huang
| Challenge: | Existing methods for achieving this alignment involve employing reinforcement learning from human feedback (RLHF) Existing approaches involve using RLHF to fine-tune LLMs based on human labels . however, RLRF is susceptible to instability during fine- tuning and presents challenges in implementation. |
| Approach: | They propose to use reinforcement learning from human feedback to fine-tune large language models with human preferences to achieve precise control of model behavior. |
| Outcome: | Experiments show that RAHF can be used to capture and manipulate representations to align with a broad spectrum of human preferences or values rather than being confined to a single concept or function. |