Towards Bridging the Gap Between Offline and Iterative Alignment via Preference Distillation
Direct preference optimization DPO is a promising offline approach for aligning large language models (LLMs) due to its simplicity, computational efficiency, and implicit modeling of human preferences. Interestingly, iterative extensions of DPO have achieved stronger performance on academic benchmarks, raising two key questions: (i) Why do iterative methods generally outperform offline ones? (ii) Can their advantages be incorporated into offline alignment? To answer the first question, our controlled experiments reveal that the explicit preference model, additionally introduced in the iterative procedure, is a key factor behind its superiority over offline methods. This insight leads us t...
| Authors | Wenbo Zhang, Wenzhuo Zhou, Hengrui Cai et al. |
|---|---|
| Primary category | cs.CL |
| Published | 7 September 2026 |
| Last updated at source | 7 September 2026 |
Silytics shows public metadata from arXiv and links to the original. Information may be incomplete or out of date. Last refreshed 17 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.