abhi @abheastu

Does this mean the work is useless? Absolutely not! I believe both of these challenges are solvable. At the very least, it indicates that there is a noninvasive signal present that can be extracted with AI, and now the right hardware setup needs to be figured out.

1 1 5 859 0

View Details

abhi @abheastu

2 weeks ago

@Reddy_XX hard to estimate with OPMs specifically, which still have a high shielding requirement there's great active shielding work from @MattBrookesMEG / @CercaMagnetics (matrix coils) but consumer-ready / no room seems unbounded

0 0 1 22 0

View Details

abhi @abheastu

2 weeks ago

cool to see this getting the attention it deserves i wrote about paths to scaling MEG a few months ago, along with comparisons to other neuroimaging modalities. link in thread!

AI at Meta @AIatMeta

2 weeks ago

We’re sharing the next major milestone in our non-invasive brain-to-text decoder research: Brain2Qwerty v2. Building on v1, which was published today in @Nature, Brain2Qwerty v2 is the highest-performing end-to-end pipeline capable of real-time sentence decoding from raw brain

664 2K 14K 5.9M 7K

3 2 16 3K 5

View Details

abhi @abheastu

2 weeks ago

writeup: axu.sh/post/an-underl…

2 0 3 182 1

View Details

abhi @abheastu

2 weeks ago

i am so sample efficient it's insane

1 0 7 306 0

View Details

AI at Meta @AIatMeta

2 weeks ago

To help accelerate neuroscience breakthroughs, we're releasing the full training code for Brain2Qwerty v1 and v2, and our partner, @bcbl_, is releasing the v1 dataset. Learn more and explore the artifacts here: go.meta.me/42ed9c

18 76 824 146K 292

View Details

abhi @abheastu

2 weeks ago

did streaming fall off or did i find purpose

2 0 9 577 0

View Details

abhi @abheastu

3 weeks ago

sleep = emotional stability, but the allure of a late night...

0 0 2 107 0

View Details

abhi @abheastu

3 weeks ago

computer, interact with my thought

0 0 2 119 0

View Details

abhi @abheastu

3 weeks ago

asking a couple silly questions about the plan before hitting build >>>>

0 0 3 97 0

View Details

Susan Zhang @suchenzang

4 weeks ago

distill for me, but not for thee!

4 12 266 39K 62

View Details

David @DavidSHolz

2 years ago

dream harder

99 304 3K 275K 179

View Details

abhi @abheastu

4 weeks ago

open source frontier intelligence is inevitable

0 1 2 342 1

View Details

abhi @abheastu

4 weeks ago

if u wanna see what post AI life looks like, come to nyc no work being done, lots of loitering

0 0 3 91 0

View Details

Nathan Lambert @natolambert

a month ago

Labs starting to pull up the ladders on the ability to diffuse AI was inevitable. Doing it without telling the user is misaligned.

NomoreID @Hangsiin

a month ago

When Fable 5 is used for frontier LLM development, it does not notify the user and instead limits the model’s capabilities through methods such as prompt modification, steering vectors, and PEFT. Anthropic estimated that this would affect approximately 0.03% of traffic.