SCALLOP: NeurIPS 2026 paper distills likelihood flow maps for few-step Boltzmann generation
jmhernandez233 · x · 2026-10-16
A paper from Cambridge, CMU, Helsinki, Meta FAIR and others accepted at NeurIPS 2026 introduces SCALLOP (SCAlable LikeLihood distillation of flOw maPs).
- Builds on F2D2's likelihood flow maps: generates samples and their log densities in few function evaluations
- Replaces F2D2's Hutchinson trace estimator with a Hutchinson-free, vectorizable likelihood distillation objective, cutting training variance and time while improving performance
- Demonstrated as a Boltzmann generator for molecular science (peptide conformations from Gaussian noise, enabling importance resampling), also validated on image datasets
Project page and paper are public.
More from Research
- Datalab details the OCR script confusion bug and cuts it 7x with a new training technique — VikParuchuri · 2026-10-16
- 'cruel' sits next to 残酷: embeddings explain VLM script confusions — VikParuchuri · 2026-10-16
- Why OCR models mix scripts: "cruel" and 残酷 have nearly identical embeddings — VikParuchuri · 2026-10-16
- OCR models keep writing the right word in the wrong alphabet — and frontier models do it too — VikParuchuri · 2026-10-16
- Tsinghua professor's ApexIntelligence raises $50M to build self-evolving foundation models — DeryaTR_ · 2026-10-16
- Anders Sandberg on fluid computation: error budgets tighten quadratically over time — anderssandberg · 2026-10-16