New paper uses determinantal point processes to sample diverse yet high-quality LLM outputs
xtimv · x · 2026-11-21
A new paper, A Model of Diverse Sampling from Language Models, tackles a familiar failure mode: ask an LLM for 10 answers and you get 10 variations of the same idea. The authors revive determinantal point processes (DPPs) — distributions over sets of outputs that favor sets whose members are individually high-quality and whose embeddings point in different directions. A striking visualization samples code for a sorting-algorithm prompt, embeds each generation with a code language model, and projects to 3D via PCA: algorithm families like Quicksort, Merge, Insertion, and Heapsort emerge as distinct clusters. The paper includes experiments showing the method works.
More from Research
- Guava distills GPT-5.4 into a 4B robot agent using 2,000 sim trajectories, zero-shot on real hardware — ChongZzZhang · 2026-11-21
- Facet-SLAM Hits 113 FPS on RTX 4090 with Triangle Splatting Real-Time SLAM — jonstephens85 · 2026-11-21
- Slay the Spire joins AI eval suites; researcher notes only Ironclad was tested — najoungkim · 2026-11-21
- RH proof author Masullo: a human came up with these ideas, I'll see it through — Singularitarian · 2026-11-21
- DPP visualized: selection probability scales with squared spanned volume — xtimv · 2026-11-21
- Deep RL agents on evidence-accumulation task mirror rodent behavioral findings — neuroecology · 2026-11-21