Moritz Helmstaedter
@mh_lab
Followers
2K
Following
124
Media
27
Statuses
278
Neuroscientist investigating the kinds of neuronal circuits you find in the main part of your brain (the cerebral cortex).
Frankfurt, Germany
Joined June 2011
What makes us human, at the neuronal network level? We discovered a dramatic (10-fold) expansion of interneuron-to-interneuron networks in the human brain compared to mouse: just published in @ScienceMagazine @MpiBrain @maxplanckpress
27
491
2K
Horrific - rockets on the Weizmann institute, a close ally with many scientist friends. Our thoughts are with you!! @WeizmannScience @maxplanckpress
0
1
12
RT @helenelab2020: Super happy that our papers #2 and #3 from our 3-person lab just went online on bioRxiv! First large-scale 3DEM dataset….
0
19
0
today's talks on diffusion models, LLMs vs children, and emergent properties in LLMs, and a final panel discussion live at .
1
1
3
Don’t miss out our panel discussion tonight featuring @MelMitchell1 @AlisonGopnik @Hsompolinsky @Uri_Hasson Thomas Hofmann. 7PM CET, 1PM Eastern, 10AM Pacific time. Live at and via
2
3
13
Starting now: Haim Sompolinsky on the representational achievements of LLMs – this and more on the Brain and LLMs in the next 24 hours at our joint @mpibrain and @leopoldina symposium. Live at and via
1
1
8
Happening today! 5.5 hours to go (1:30 PM CET). Live stream at and via
0
4
6
Let’s hope this works out! . Live stream will be available, and the panel discussions will stay online. 6/end.
0
0
1
and what this tells us, from a scientific standpoint, about the expected trajectory of today's AI architectures. A main ambition is: to employ scientific discourse, rather than invectives obfuscated by company interests, for diving deep into these topics. 5/.
1
0
0
whether language is indeed a learning bottleneck, if mortal intelligence is indeed doomed, what compressed vs. uncompressed data can achieve during training, what the fundamental limitations of current network architectures may be – 4/.
1
0
0
Talks span from the analysis of the geometry of language embeddings to the phenotyping of LLM “behavior” using methods from child psychology. And we will probe what experts think about “understanding” in LLMs, the validity of examples vs counterexamples of reasoning 3/.
1
0
0
@mtoneva1 @Hsompolinsky @Uri_Hasson @MelMitchell1 @AlisonGopnik @eunice_yiu_ @Igurevych @kerstingAIML @ml_tuberlin @ViolaPriesemann @gemmarono @wawiese Björn Ommer and Thomas Hofmann. Co-organizers Thomas Lengauer and Johannes Buchmann. 2/.
1
1
3
Can Brain Science help us understand what Large Language Models have achieved? Thrilled to announce our symposium May 13+14 organized jointly by @mpibrain and @leopoldina . Live stream available via Exciting talks and panel discussion with 1/
2
11
26
RT @SebastianSeung: @kendmil Ah a connectomic answer to the question "What makes us human?" Or at least "What makes us primates?" Though he….
0
1
0
RT @FENSorg: 🏆 #Congratulations to The @BrainPrize winners: Larry Abbott, @HSompolinsky and @sejnowski!. 🧬 This #prize recognises their pio….
0
28
0
RT @helenelab2020: Every synapse counts!! First 3D EM-connectomic circuit analysis of mammalian memory networks. And first paper from the l….
0
11
0
