Reasoning VLAs can think. They just can't think fast. Until now.
@zhijianliu10K followers
192,223 views
32 Comments
Comments from the original post on X (Twitter)

@zhijianliu_
Code + model checkpoints coming soon. The ingredients are already open-source: 🔥 DFlash → ⚡ ParoQuant →
35 likes

@zhijianliu_
Great work led by @Richard91316073 and @YihaoLiang01, with @hongfeizhang0xF and @jianchen1799.
13 likes

@idontev63124802
Thanks for the paper, I've spent around 2 hours reading it and come up with my response. Screenshot because of twitter character limit.
13 likes

@zhijianliu_
Thanks for sharing your feedback! The baseline is under BF16.
5 likes

@tugot17
awesome work 🫡
5 likes

@crepesupreme
716ms to 159ms is the line real-time robotics has been waiting on. Every VLA roadmap priced in at least another year before reasoning models could sit inside the control loop. Quietly huge.
4 likes

@atris_eth
Trying this on an in-car 3090 with a comma soon
3 likes

@MaoLi92351
5.7x faster with zero accuracy loss sounds like the kind of magic trick that only works on paper until you actually try it in a car moving at 60mph
3 likes

@thehenryinsf
holy shit. need this deployed somewhere and served on an api right now
3 likes

@donpark
could be more compelling from safety perspective if the model used for speculative reasoning is RL-ed to be more cautious than the full reasoning. safety-first in the literal sense. :-)
2 likes

@brianchew
@kstonekuan this looks interesting and in your field!
2 likes

@techietaro
Sub-160ms VLAs that don't think in molasses. ParoQuant + DFlash cuts latency 5.7x with zero accuracy loss. Lab demos just became production reality. ⚡️
2 likes

@bronzeagepapi
@yassineyousfi_
2 likes

@avaisaziz
Nice work getting reasoning VLA down to 159ms on RTX PRO 6000 with no accuracy drop. Streaming + DFlash speculative decoding + ParoQuant W4A8 is a clean stack for closing the control loop. Looking forward to the checkpoints.
2 likes

@AlenGabs
very cool !
1 like

@Reagent_Systems
Holy smokes this is really cool :O
1 like

@uguryektabasak
this is very practical
1 like

@AiAtlas7555
716ms to 159ms. Real-time VLA reasoning is no longer a lab benchmark.
1 like

@NischayJoshi8
The latency gap is real for live video reasoning. The quality tradeoff at high token budgets is the other side - we found diminishing returns past ~700 thinking tokens for video understanding. Benchmark at
1 like

@davidzhaiyang
Really cool!
1 like








