mlsanityData to 14 Oct 2026

Unsupervised Post-Training of Foundation Models: A Survey

arXiv 2608.24982 ↗First posted 25 Aug 2026cs.CL cs.AI cs.CVcited by 0

Not on the map. Why?

Abstract (arXiv)

Foundation-model post-training usually relies on human labels, preference data, stronger teachers, or executable verifiers. We study Unsupervised Post-Training (UPT): update-bearing adaptation on unlabeled inputs whose learning signal is derived from same-lineage model artifacts rather than an external oracle. We catalog 80 strict UPT methods and organize them by the object that supplies the update signal: a prediction statistic, a sample relation, a self-generated target, or an internal evaluator. Beyond inventory, we show how the choice of internal signal and task structure determines whether post-training improves the model or recursively amplifies error. An orthogonal Input Visibility × Update Persistence view maps deployment regimes and defines a unified framework for UPT selection and evaluation.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Builds on 69 of its 87 references that are in the map. First posted on arXiv 25 Aug 2026.

Builds on

69 of 87 references.

References · 69

  1. 2024DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Modelscited by 9,379
  2. 2022Self-Instruct: Aligning Language Models with Self-Generated Instructionscited by 2,724
  3. 2022Large Language Models Can Self-Improvecited by 664
  4. 2019Pseudo-Labeling and Confirmation Bias in Deep Semi-Supervised Learningcited by 510
  5. 2023A Comprehensive Survey on Test-Time Adaptation under Distribution Shiftscited by 359
  6. 2025Absolute Zero: Reinforced Self-play Reasoning with Zero Datacited by 344
  7. 2024Quiet-STaR: Language Models Can Teach Themselves to Think Before Speakingcited by 301
  8. 2025R-Zero: Self-Evolving Reasoning LLM from Zero Datacited by 228
  9. 2025Learning to Reason without External Rewardscited by 210
  10. 2024Meta-Rewarding Language Models: Self-Improving Alignment with LLM-as-a-Meta-Judgecited by 204
  11. 2023Continual Pre-training of Language Modelscited by 198
  12. 2024Data Engineering for Scaling Language Models to 128K Contextcited by 186
38 more
  1. 2020Theoretical Analysis of Self-Training with Deep Networks on Unlabeled Datacited by 180
  2. 2025The Unreasonable Effectiveness of Entropy Minimization in LLM Reasoningcited by 166
  3. 2025Right Question is Already Half the Answer: Fully Unsupervised LLM Reasoning Incentivizationcited by 128
  4. 2025Maximizing Confidence Alone Improves Reasoningcited by 100
  5. 2025Self-Training Elicits Concise Reasoning in Large Language Modelscited by 99
  6. 2025Can Large Reasoning Models Self-Train?cited by 68
  7. 2025Confidence Is All You Need: Few-Shot RL Fine-Tuning of Language Modelscited by 62
  8. 2025Consistent Paths Lead to Truth: Self-Rewarding Reinforcement Learning for LLM Reasoningcited by 46
  9. 2025Evolving Language Models without Labels: Majority Drives Selection, Novelty Promotes Variationcited by 45
  10. 2025ETTRL: Balancing Exploration and Exploitation in LLM Test-Time Reinforcement Learning Via Entropy Mechanismcited by 36
  11. 2024CREAM: Consistency Regularized Self-Rewarding Language Modelscited by 35
  12. 2025First SFT, Second RL, Third UPT: Continual Improving Multi-Modal LLM Reasoning via Unsupervised Post-Trainingcited by 34
  13. 2024Large Language Models Can Self-Improve in Long-context Reasoningcited by 29
  14. 2025EvoLMM: Self-Evolving Large Multimodal Models with Continuous Rewardscited by 27
  15. 2025LongPO: Long Context Self-Evolution of Large Language Models through Short-to-Long Preference Optimizationcited by 24
  16. 2025SePer: Measure Retrieval Utility Through The Lens Of Semantic Perplexity Reductioncited by 24
  17. 2025Genius: A Generalizable and Purely Unsupervised Self-Training Framework For Advanced Reasoningcited by 24
  18. 2026TTCS: Test-Time Curriculum Synthesis for Self-Evolvingcited by 19
  19. 2025SLOT: Sample-specific Language Model Optimization at Test-timecited by 18
  20. 2025Let's (not) just put things in Context: Test-Time Training for Long-Context LLMscited by 18
  21. 2024Self-Tuning: Instructing LLMs to Effectively Acquire New Knowledge through Self-Teachingcited by 17
  22. 2026G-Zero: Self-Play for Open-Ended Generation from Zero Datacited by 17
  23. 2025Sailing by the Stars: A Survey on Reward Models and Learning Strategies for Learning from Rewardscited by 14
  24. 2026On-the-Fly VLA Adaptation via Test-Time Reinforcement Learningcited by 14
  25. 2025Beyond Majority Voting: Towards Fine-grained and More Reliable Reward Signal for Test-Time Reinforcement Learningcited by 13
  26. 2025SPINE: Token-Selective Test-Time Reinforcement Learning with Entropy-Band Regularizationcited by 12
  27. 2026What If Consensus Lies? Selective-Complementary Reinforcement Learning at Test Timecited by 12
  28. 2025Self-Evolving Vision-Language Models for Image Quality Assessment via Voting and Rankingcited by 10
  29. 2025Self-Harmony: Learning to Harmonize Self-Supervision and Self-Play in Test-Time Reinforcement Learningcited by 10
  30. 2026Tool Verification for Test-Time Reinforcement Learningcited by 9
  31. 2025Self-Training Large Language Models with Confident Reasoningcited by 8
  32. 2025SUDER: Self-Improving Unified Large Multimodal Models for Understanding and Generation with Dual Self-Rewardscited by 8
  33. 2025Self-Improvement of Language Models by Post-Training on Multi-Agent Debatecited by 7
  34. 2025Model Whisper: Steering Vectors Unlock Large Language Models' Potential in Test-timecited by 7
  35. 2026Silence the Judge: Reinforcement Learning with Self-Verifier via Latent Geometric Clusteringcited by 7
  36. 2025DEBATE, TRAIN, EVOLVE: Self Evolution of Language Model Reasoningcited by 6
  37. 2025LongMagpie: A Self-synthesis Method for Generating Large-scale Long-context Instructionscited by 6
  38. 2025Wisdom of the Crowd: Reinforcement Learning from Coevolutionary Collective Feedbackcited by 6

50 most cited of 69 All 69 (JSON)

Built on by

Not cited yet.