# Curry-DPO: Enhancing Alignment using Curriculum Learning & Ranked Preferences

arXiv: https://arxiv.org/abs/2403.07230 · HTML: https://mlsanity.com/paper/2403.07230/

## Key facts

- Cited by 39 papers in the map (data through 14 Oct 2026).
- Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 25 of its 51 references that are in the map.
- First posted on arXiv 12 Mar 2024.

## Placed in

- LLM alignment / RLHF / preference optimisation / [LLM preference optimization methods](https://mlsanity.com/alignment/llm-preference-optimization-methods/)

## Builds on (most cited first, 25 of 25)

- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [Llama 2: Open Foundation and Fine-Tuned Chat Models](https://mlsanity.com/paper/2307.09288/) (2023)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Deep reinforcement learning from human preferences](https://mlsanity.com/paper/1706.03741/) (2017)
- [Training a Helpful and Harmless Assistant with Reinforcement Learning from Human Feedback](https://mlsanity.com/paper/2204.05862/) (2022)
- [Learning to summarize from human feedback](https://mlsanity.com/paper/2009.01325/) (2020)
- [A General Theoretical Paradigm to Understand Learning from Human Preferences](https://mlsanity.com/paper/2310.12036/) (2023)
- [Enhancing Chat Language Models by Scaling High-quality Instructional Conversations](https://mlsanity.com/paper/2305.14233/) (2023)
- [OpenAssistant Conversations -- Democratizing Large Language Model Alignment](https://mlsanity.com/paper/2304.07327/) (2023)
- [Self-Play Fine-Tuning Converts Weak Language Models to Strong Language Models](https://mlsanity.com/paper/2401.01335/) (2024)
- [RRHF: Rank Responses to Align Language Models with Human Feedback without tears](https://mlsanity.com/paper/2304.05302/) (2023)
- [Catastrophic Jailbreak of Open-source LLMs via Exploiting Generation](https://mlsanity.com/paper/2310.06987/) (2023)
- [SLiC-HF: Sequence Likelihood Calibration with Human Feedback](https://mlsanity.com/paper/2305.10425/) (2023)
- [Competence-based Curriculum Learning for Neural Machine Translation](https://mlsanity.com/paper/1903.09848/) (2019)
- [ProsocialDialog: A Prosocial Backbone for Conversational Agents](https://mlsanity.com/paper/2205.12688/) (2022)
- [Some things are more CRINGE than others: Iterative Preference Optimization with the Pairwise Cringe Loss](https://mlsanity.com/paper/2312.16682/) (2023)
- [LiPO: Listwise Preference Optimization through Learning-to-Rank](https://mlsanity.com/paper/2402.01878/) (2024)
- [Can Neural Machine Translation be Improved with User Feedback?](https://mlsanity.com/paper/1804.05958/) (2018)
- [Curriculum Learning for Domain Adaptation in Neural Machine Translation](https://mlsanity.com/paper/1905.05816/) (2019)
- [Simple and Effective Curriculum Pointer-Generator Networks for Reading Comprehension over Long Narratives](https://mlsanity.com/paper/1905.10847/) (2019)
- [Step-On-Feet Tuning: Scaling Self-Alignment of LLMs via Bootstrapping](https://mlsanity.com/paper/2402.07610/) (2024)
- [Exploiting Curriculum Learning in Unsupervised Neural Machine Translation](https://mlsanity.com/paper/2109.11177/) (2021)

## Built on by (most cited first, 30 of 39)

- [DUMP: Automated Distribution-Level Curriculum Learning for RL-based LLM Post-training](https://mlsanity.com/paper/2504.09710/) (2025)
- [Principled Data Selection for Alignment: The Hidden Risks of Difficult Examples](https://mlsanity.com/paper/2502.09650/) (2025)
- [A Survey of Direct Preference Optimization](https://mlsanity.com/paper/2503.11701/) (2025)
- [ExCoT: Optimizing Reasoning for Text-to-SQL with Execution Feedback](https://mlsanity.com/paper/2503.19988/) (2025)
- [BPO: Towards Balanced Preference Optimization between Knowledge Breadth and Depth in Alignment](https://mlsanity.com/paper/2411.10914/) (2024)
- [Geometric-Averaged Preference Optimization for Soft Preference Labels](https://mlsanity.com/paper/2409.06691/) (2024)
- [What Matters in Data for DPO?](https://mlsanity.com/paper/2508.18312/) (2025)
- [Larger or Smaller Reward Margins to Select Preferences for Alignment?](https://mlsanity.com/paper/2503.01864/) (2025)
- [Hummer: Towards Limited Competitive Preference Dataset](https://mlsanity.com/paper/2405.11647/) (2024)
- [Multi-Reference Preference Optimization for Large Language Models](https://mlsanity.com/paper/2405.16388/) (2024)
- [Value Drifts: Tracing Value Alignment During LLM Post-Training](https://mlsanity.com/paper/2510.26707/) (2025)
- [Self-Augmented Preference Optimization: Off-Policy Paradigms for Language Model Alignment](https://mlsanity.com/paper/2405.20830/) (2024)
- [Adaptive Helpfulness-Harmlessness Alignment with Preference Vectors](https://mlsanity.com/paper/2504.20106/) (2025)
- [LaMDAgent: An Autonomous Framework for Post-Training Pipeline Optimization via LLM Agents](https://mlsanity.com/paper/2505.21963/) (2025)
- [VDC-Agent: When Video Detailed Captioners Evolve Themselves via Agentic Self-Reflection](https://mlsanity.com/paper/2511.19436/) (2025)
- [SeRA: Self-Reviewing and Alignment of Large Language Models using Implicit Reward Margins](https://mlsanity.com/paper/2410.09362/) (2024)
- [REWARD CONSISTENCY: Improving Multi-Objective Alignment from a Data-Centric Perspective](https://mlsanity.com/paper/2504.11337/) (2025)
- [FRAME: Boosting LLMs with A Four-Quadrant Multi-Stage Pretraining Strategy](https://mlsanity.com/paper/2502.05551/) (2025)
- [Towards Understanding Valuable Preference Data for Large Language Model Alignment](https://mlsanity.com/paper/2510.13212/) (2025)
- [Teaching with Lies: Curriculum DPO on Synthetic Negatives for Hallucination Detection](https://mlsanity.com/paper/2505.17558/) (2025)
- [A Deep Dive into Scaling RL for Code Generation with Synthetic Data and Curricula](https://mlsanity.com/paper/2603.24202/) (2026)
- [Understanding the Impact of Sampling Quality in Direct Preference Optimization](https://mlsanity.com/paper/2506.04272/) (2025)
- [From Fragments to Facts: A Curriculum-Driven DPO Approach for Generating Hindi News Veracity Explanations](https://mlsanity.com/paper/2507.05179/) (2025)
- [Improving LLMs for Machine Translation Using Synthetic Preference Data](https://mlsanity.com/paper/2508.14951/) (2025)
- [On the Transformations across Reward Model, Parameter Update, and In-Context Prompt](https://mlsanity.com/paper/2406.16377/) (2024)
- [Learning More with Less: A Dynamic Dual-Level Down-Sampling Framework for Efficient Policy Optimization](https://mlsanity.com/paper/2509.22115/) (2025)
- [Fine-Tuning LLMs with Fine-Grained Human Feedback on Text Spans](https://mlsanity.com/paper/2512.23693/) (2025)
- [CLewR: Curriculum Learning with Restarts for Machine Translation Preference Learning](https://mlsanity.com/paper/2601.05858/) (2026)
- [Entropy-Guided Data-Efficient Training for Multimodal Reasoning Reward Models](https://mlsanity.com/paper/2602.01884/) (2026)
- [Curriculum Group Policy Optimization: Adaptive Sampling for Unleashing the Potential of Text-to-Image Generation](https://mlsanity.com/paper/2605.17807/) (2026)

## Abstract (arXiv)

Direct Preference Optimization (DPO) is an effective technique that leverages pairwise preference data (usually one chosen and rejected response pair per user prompt) to align LLMs to human preferences. In practice, multiple responses can exist for a given prompt with varying quality relative to each other. With availability of such quality ratings for multiple responses, we propose utilizing these responses to create multiple preference pairs for a given prompt. Our work focuses on systematically using the constructed multiple preference pair in DPO training via curriculum learning methodology. In particular, we order these multiple pairs of preference data from easy to hard (emulating curriculum training) according to various criteria. We show detailed comparisons of our proposed approach to the standard single-pair DPO setting. Our method, which we call Curry-DPO consistently shows increased performance gains on MTbench, Vicuna, WizardLM, and the UltraFeedback test set, highlighting its effectiveness. More specifically, Curry-DPO achieves a score of 7.43 on MT-bench with Zephy-7B model outperforming majority of existing LLMs with similar parameter size. Curry-DPO also achieves the highest adjusted win rates on Vicuna, WizardLM, and UltraFeedback test datasets (90.7%, 87.1%, and 87.9% respectively) in our experiments, with notable gains of upto 7.5% when compared to standard DPO technique. We release the preference pairs used in alignment at: https://huggingface.co/datasets/ServiceNow-AI/Curriculum_DPO_preferences