Quyet V. Do

@Quyet_Azir

PhD at @virginia_tech, supervised by @tuvllms. Work on LLMs' long-context search, post-training, and self evolution. ex-Research Intern at @AdobeResearch.

Blacksburg, VA

sites.google.com/view/williamqu…

Joined May 2022

406 Following

141 Followers

Replies Reposts Media

Pinned
Quyet V. Do
@Quyet_Azir
Oct 4
🚀 I am excited to share our recent work, which I will present at #COLM2026 🚨 Our data curation framework improves LLMs’ long-context reasoning capability in open domains, even via self-distillation! We introduce Pi^2, a pipeline that 1) transforms Wikipedia tables into
2
Quyet V. Do
@Quyet_Azir
Nov 3
What a fascinating emergent capability of frontier models! It would definitely be helpful for recursive self-improvement (RSI).
Rikiya Takehi
@rikiyatakehi
Nov 2
We asked GPT-6 Astra to edit Qwen3.5-0.8B weights to improve math. It directly changed only 0.016% of weights, fixing 0.8B's reasoning loops & improving math up to 6x and even instruction following 3x, without gradients or training. We call this Weight-Space Programming:
Readers added context
LLM was given weights explicitly as text in prompt and used code/forward passes (per authors' logs) to ablate Qwen looping direction, a known thinking-mode issue per model card. Primarily fixed repetition in that mode, not general 'natively' improving weights/math. rikiyat.github.io/blog/weight-sp… github.com/RikiyaT/lm_und… huggingface.co/Qwen/Qwen3.5-0… huggingface.co/RikiyaT/Qwen3.…
Quyet V. Do
@Quyet_Azir
Oct 6
I’ll be presenting my #COLM2026 paper today 👋 📜 Pi^2: Structure-Originated Reasoning Data Improves Long-Context Reasoning Ability of Large Language Models 📍Location: Poster Session 1 (11am-1pm), #134, Grand Ballroom. 🌐︎ Virtual site: colm.eventhosts.cc/virtual/2026/p… If you are
Quyet V. Do
@Quyet_Azir
Jun 20, 2025
Wander around @Adobe with @reds_tiger and see a Macintosh in real life for the first time 😮
Quyet V. Do
@Quyet_Azir
Mar 26, 2025
Congrats @linusdd44804 🎉 Very interesting paper!
Tu Vu
@tuvllms
Mar 26, 2025
🚨 New paper 🚨 Excited to share my first paper w/ my PhD students!! We find that advanced LLM capabilities conferred by instruction or alignment tuning (e.g., SFT, RLHF, DPO, GRPO) can be encoded into model diff vectors (à la task vectors) and transferred across model
1

Quyet V. Do

Quyet V. Do’s posts