Discrete Diffusion Reading Group

@diffusion_llms

📚 Journal club on discrete diffusion models 🎥 Replays available on YouTube! Contact: [email protected] Hosted by @ssahoo_, @jdeschena, @zhihanyang_

d-llms.com

Joined August 2025

Following

2,304

Followers

Replies Reposts Media

Pinned
Discrete Diffusion Reading Group
@diffusion_llms
Oct 30, 2025
Drowning in the sea of Discrete Diffusion papers? 🌊 We got you. Join our Reading Group! From theory → empirics, and language → molecules — we’ll decode the chaos together 💫 Join the cult—uh, I mean community 😇 👉 Google Group: groups.google.com/g/diffusion-ll… (1 / 2)
2
Discrete Diffusion Reading Group
@diffusion_llms
Sep 15
This week, @ssahoo_ presents diffusion-augmented LLMs: 🔥 AR-level Quality ⚡️ Diffusion-level speed ✨ A drop-in replacement for speculative decoding: >> No separate draft model to train, plus >> Faster RL post-training. The method, Uno, beats all diffusion
Subham Sahoo
@ssahoo_
Sep 4
Diffusion LLMs have two limitations relative to AR models: (1) Lower quality, and (2) Slower inference at large batch sizes. We address this "Uno" > Retains the AR architecture of LLMs > Each layer has two sets of weights: AR weights and Diffusion weights > Diffusion
1
Discrete Diffusion Reading Group
@diffusion_llms
Sep 7
Missed the talk? Check out the recording on YouTube: youtube.com/watch?v=jfUyPG…
Discrete Diffusion Reading Group
@diffusion_llms
Sep 4
🤔 Can masked models get deeper without getting bigger? Three recent papers converge on the same insight from different directions: reuse computation instead of adding parameters. Depth is normally fixed at training time — these works make it something you can spend at
Discrete Diffusion Reading Group
@diffusion_llms
Sep 4
🤔 Can masked models get deeper without getting bigger? Three recent papers converge on the same insight from different directions: reuse computation instead of adding parameters. Depth is normally fixed at training time — these works make it something you can spend at
4
Discrete Diffusion Reading Group
@diffusion_llms
Sep 4
Discrete Diffusion goes mainstream!! 🎊
Subham Sahoo
@ssahoo_
Sep 4
Diffusion LLMs have two limitations relative to AR models: (1) Lower quality, and (2) Slower inference at large batch sizes. We address this "Uno" > Retains the AR architecture of LLMs > Each layer has two sets of weights: AR weights and Diffusion weights > Diffusion