Search papers, labs, and topics across Lattice.
KAIST
4
0
4
Fixed-point flows enable a leap in performance for language models, outperforming state-of-the-art methods in one- and few-step generation tasks.
Achieving high-fidelity language generation with 32x fewer function evaluations could revolutionize real-time applications of language models.
Flow-based generative models can be improved *after* training, simply by steering intermediate states away from regions where the velocity field is highly divergent.
Forget slow, multi-step diffusion: this work achieves state-of-the-art text generation quality with a *single* denoising step using flow-based language models.
