Session 19: ELF: Embedded Language Flows

[Diffusion LLM]

Hello folks,

Unlike their image-domain counterparts, today’s leading diffusion language models (DLMs) primarily operate over discrete tokens. 

The authors of ELF show that continuous DLMs can be made effective with minimal adaptation to the discrete domain. They propose Embedded Language Flows (ELF), a class of diffusion models in continuous embedding space based on continuous-time Flow Matching. Unlike existing DLMs, ELF predominantly stays within the continuous embedding space until the final time step, where it maps to discrete tokens using a shared-weight network. 

This formulation makes it straightforward to adapt established techniques from image-domain diffusion models, e.g., classifier-free guidance (CFG). 

Experiments show that ELF substantially outperforms leading discrete and continuous DLMs, achieving better generation quality with fewer sampling steps. These results suggest that ELF offers a promising path toward effective continuous DLMs.

This Monday, Keya Hu and Linlu Qiu will present their jointly led paper ELF.

Title: ELF: Embedded Language Flows

Meeting Link: click here

Time: June 1 (Monday) 1pm ET / 10am PT / 7pm CET / 10:30pm IST

Paper: [2605.10938] ELF: Embedded Language Flows

Prior knowledge: 

Fundamentals of discrete diffusion (video by Sasha Rush)

The Diffusion Duality (video by our reading group)

Abstract:

Diffusion and flow-based models have become the de facto approaches for generating continuous data, e.g., in domains such as images and videos. Their success has attracted growing interest in applying them to language modeling. Unlike their image-domain counterparts, today's leading diffusion language models (DLMs) primarily operate over discrete tokens. In this paper, we show that continuous DLMs can be made effective with minimal adaptation to the discrete domain. We propose Embedded Language Flows (ELF), a class of diffusion models in continuous embedding space based on continuous-time Flow Matching. Unlike existing DLMs, ELF predominantly stays within the continuous embedding space until the final time step, where it maps to discrete tokens using a shared-weight network. This formulation makes it straightforward to adapt established techniques from image-domain diffusion models, e.g., classifier-free guidance (CFG). Experiments show that ELF substantially outperforms leading discrete and continuous DLMs, achieving better generation quality with fewer sampling steps. These results suggest that ELF offers a promising path toward effective continuous DLMs.

Yours truly,

Subham, Justin, Zhihan

Website, Twitter, Discord, YouTube