Grace Luo

@graceluo_

phd student @berkeley_ai, vision + language

graceluo.net

Joined June 2021

Replies Reposts Media

Pinned
Grace Luo
@graceluo_
Feb 9
We trained diffusion models on a billion LLM activations, and we want you to use them! New preprint: Learning a Generative Meta-Model of LLM Activations Joint work with @feng_jiahai, @trevordarrell, @AlecRad, @JacobSteinhardt. More in thread 🧵
00:00
32
Grace Luo
@graceluo_
Jul 8
I’m at #ICML2026 in Seoul and presenting a poster *today* in the morning session! Come nerd out with me about training diffusion models on LLM activations (or swing by and say hi!). 📌 icml.cc/virtual/2026/p… 🔗 generative-latent-prior.github.io
1
Grace Luo
@graceluo_
Jul 7
I’m at #ICML2026 and presenting a poster *today* in the morning session! Come nerd out with me about training diffusion models on LLM activations (or swing by and say hi!). 📌 icml.cc/virtual/2026/p… 🔗 generative-latent-prior.github.io
1
Grace Luo
@graceluo_
Oct 22, 2025
If you're at #ICCV2025, swing by our poster today! We'll show how you can directly fine-tune an image generator with a VLM's language modeling loss. 📍iccv.thecvf.com/virtual/2025/p… 🔗 dual-process.github.io cc @jongranskog @holynski_ @trevordarrell
1
Grace Luo
@graceluo_
Oct 20, 2025
Check out our new project at echo-bench.github.io! Go and poke the visualizations on our website; you might even find a few easter eggs 🐣
Jiaxin Ge
@jiaxin_ge_
Oct 20, 2025
✨Introducing ECHO, the newest in-the-wild image generation benchmark! You’ve seen new image models and new use cases discussed on social media, but old benchmarks don’t test them! We distilled this qualitative discussion into a structured benchmark. 🔗 echo-bench.github.io
00:00