# LLM alignment / RLHF / preference optimisation

5,228 papers in 10 research directions. Data through 7 Oct 2026.
HTML: https://mlsanity.com/alignment/

## Directions (largest first)

- [LLM preference optimization methods](https://mlsanity.com/alignment/llm-preference-optimization-methods/index.md): 1,132 papers, 3 new this week
- [Reward modeling and RLHF](https://mlsanity.com/alignment/reward-modeling-and-rlhf/index.md): 1,029 papers, 2 new this week
- [Human values and AI feedback alignment](https://mlsanity.com/alignment/human-values-and-ai-feedback-alignment/index.md): 709 papers, 2 new this week
- [Diffusion and multimodal model alignment](https://mlsanity.com/alignment/diffusion-and-multimodal-model-alignment/index.md): 709 papers, 2 new this week
- [Safety alignment and jailbreak attacks](https://mlsanity.com/alignment/safety-alignment-and-jailbreak-attacks/index.md): 687 papers, 7 new this week
- [Process reward models for reasoning](https://mlsanity.com/alignment/process-reward-models-for-reasoning/index.md): 576 papers, 4 new this week
- [Preference alignment for recommenders and RAG](https://mlsanity.com/alignment/preference-alignment-for-recommenders-and-rag/index.md): 169 papers, 0 new this week
- [Medical LLM alignment](https://mlsanity.com/alignment/medical-llm-alignment/index.md): 104 papers, 2 new this week
- [Reward hacking and misspecification](https://mlsanity.com/alignment/reward-hacking-and-misspecification/index.md): 94 papers, 2 new this week
- [Preference optimization for molecular design](https://mlsanity.com/alignment/preference-optimization-for-molecular-design/index.md): 19 papers, 0 new this week