mlsanityData to 7 Oct 2026

WILDCHAT-50M: A Deep Dive Into the Role of Synthetic Data in Post-Training

arXiv 2501.18511 ↗First posted 30 Jan 2025cs.LG cs.CLcited by 4

On the map: LLM alignment / RLHF / preference optimisation → LLM preference optimization methods

also in 1 more
Abstract (arXiv)

Language model (LLM) post-training, from DPO to distillation, can refine behaviors and unlock new skills, but the open science supporting these post-training techniques is still in its infancy. One limiting factor has been the difficulty of conducting large-scale comparative analyses of synthetic data generating models and LLM judges. To close this gap, we introduce WILDCHAT-50M, the largest public chat dataset to date. We extend the existing WildChat dataset to include responses not only from GPT, but from over 50 different open-weight models, ranging in size from 0.5B to 104B parameters. We conduct an extensive comparative analysis and demonstrate the potential of this dataset by creating RE-WILD, our own public SFT mix, which outperforms the recent Tulu-3 SFT mixture from Allen AI with only 40% as many samples. Our dataset, samples and code are available at https://github.com/penfever/wildchat-50m.

Key facts. Cited by 4 papers in the map (data through 7 Oct 2026). Placed in LLM preference optimization methods, one of 10 directions in LLM alignment / RLHF / preference optimisation. Builds on 33 of its 53 references that are in the map. First posted on arXiv 30 Jan 2025.

Builds on

33 of 53 references.

Closest ancestors · 5

Same direction, newest first

  1. 2024Magpie: Alignment Data Synthesis from Scratch by Prompting Aligned LLMs with Nothing12 Jun
  2. 2024SimPO: Simple Preference Optimization with a Reference-Free Reward23 May
  3. 2023Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 217 Nov
  4. 2023Zephyr: Direct Distillation of LM Alignment25 Oct
  5. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructions20 Dec

Other references · 28

  1. 2024The Llama 3 Herd of Modelscited by 14,192
  2. 2023Llama 2: Open Foundation and Fine-Tuned Chat Modelscited by 12,865
  3. 2023Direct Preference Optimization: Your Language Model is Secretly a Reward Modelcited by 7,916
  4. 2020Measuring Massive Multitask Language Understandingcited by 7,768
  5. 2023Efficient Memory Management for Large Language Model Serving with PagedAttentioncited by 5,767
  6. 2021Finetuned Language Models Are Zero-Shot Learnerscited by 3,747
  7. 2023Mistral 7Bcited by 3,065
  8. 2017Decoupled Weight Decay Regularizationcited by 1,722
  9. 2021Multitask Prompted Training Enables Zero-Shot Task Generalizationcited by 1,462
  10. 2024Chatbot Arena: An Open Platform for Evaluating LLMs by Human Preferencecited by 1,151
  11. 2023WizardLM: Empowering large pre-trained language models to follow complex instructionscited by 1,116
  12. 2024ChatGLM: A Family of Large Language Models from GLM-130B to GLM-4 All Toolscited by 1,096
16 more
  1. 2020On Faithfulness and Factuality in Abstractive Summarizationcited by 895
  2. 2024Length-Controlled AlpacaEval: A Simple Way to Debias Automatic Evaluatorscited by 865
  3. 2022Super-NaturalInstructions: Generalization via Declarative Instructions on 1600+ NLP Taskscited by 858
  4. 2023The Flan Collection: Designing Data and Methods for Effective Instruction Tuningcited by 734
  5. 2021Cross-Task Generalization via Natural Language Crowdsourcing Instructionscited by 696
  6. 2024WildChat: 1M ChatGPT Interaction Logs in the Wildcited by 598
  7. 2023How Far Can Camels Go? Exploring the State of Instruction Tuning on Open Resourcescited by 405
  8. 2024Preference Fine-Tuning of LLMs Should Leverage Suboptimal, On-Policy Datacited by 184
  9. 2024Nemotron-4 340B Technical Reportcited by 143
  10. 2024MixEval: Deriving Wisdom of the Crowd from LLM Benchmark Mixturescited by 73
  11. 2018ROUGE 2.0: Updated and Improved Measures for Evaluation of Summarization Taskscited by 66
  12. 2024A Comprehensive Evaluation of Quantization Strategies for Large Language Modelscited by 47
  13. 2024Style Outweighs Substance: Failure Modes of LLM Judges in Alignment Benchmarkingcited by 35
  14. 2024Hermes 3 Technical Reportcited by 33
  15. 2023Sparse Fine-tuning for Inference Acceleration of Large Language Modelscited by 14
  16. 2024SELECT: A Large-Scale Benchmark of Data Curation Strategies for Image Classificationcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2025: 2; 2026: 2.

4 papers.

  1. 2025MemLoRA: Distilling Expert Adapters for On-Device Memory Systemscited by 7
  2. 2026Beyond LLM-as-a-Judge: Deterministic Metrics for Multilingual Generative Text Evaluationcited by 3
  3. 2025LLM-Driven Preference Data Synthesis for Proactive Prediction of the Next User Utterance in Human-Machine Dialoguecited by 2
  4. 2026CompanionSim: Synthetic Data for Evaluating Anthropomorphism in Human-AI Relationshipscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2023Chat Vector: A Simple Approach to Equip LLMs with Instruction Following and Model Alignment in New Languages
  2. 2023A Systematic Study and Comprehensive Evaluation of ChatGPT on Benchmark Datasets
  3. 2024On the Diversity of Synthetic Data and its Impact on Training Large Language Models
  4. 2025Demystifying Synthetic Data in LLM Pre-training: A Systematic Study of Scaling Laws, Benefits, and Pitfalls
  5. 2024ChatGPT Alternative Solutions: Large Language Models Survey
  6. 2024Hybrid Training Approaches for LLMs: Leveraging Real and Synthetic Data to Enhance Model Performance in Domain-Specific Applications
  7. 2023Summary of ChatGPT-Related Research and Perspective Towards the Future of Large Language Models
  8. 2026It's All Training: A Fully Synthetic Single-Stage Recipe for LLMs