Deft model fixes LLM formulaic writing with Distribution Fine-Tuning
HamelHusain · x · 2026-08-20
Deft is a new model and lab focused on improving AI writing. Its research paper highlights that standard SFT leads to formulaic outputs, overusing certain tokens or phrases ("slop signs").
Deft introduces Distribution Fine-Tuning (DFT), a post-training step that aligns the model's output distribution more closely with the training data. Experiments show DFT improves MMD by 49% and JMQ by 63%. The model significantly boosts creativity, coherence, and clarity, achieving a 100% human-written score on the Pangram detector.
More from Research
- Ising model explains consensus and polarization in 10k LLM agents — bronzeagepapi · 2026-08-20
- Recirculation: Off-the-shelf models self-modify for instant inference boost — TheGradient · 2026-08-20
- Describing multi-agent interactions as a dynamical system to predict consensus — fatihdin4en · 2026-08-20
- RAG poisoning creates false confidence: monitoring attention beats uncertainty checks — rohanpaul_ai · 2026-08-20
- Weak adversarial networks solve NS-Brinkman equations, outperforming PINNs — FrnkNlsn · 2026-08-20
- Stanford releases quickstart guide for ENCODE GRAMMAR genomics AI tool — anshulkundaje · 2026-08-20