Encode Fellow of the 2025 cohort. PhD from University of Oxford botcs.github.io/academic
Joined May 2017
- Tweets 330
- Following 310
- Followers 264
- Likes 1,781
71 Photos and videos
Pinned Tweet
May 11
Jeee 🐦⬛
I am very proud of our joint effort with @sreejan_kumar Kumar on the project "Reason to Play"
LRMs show human-like rule discovery, and their hidden states predict human brain activity during gameplay.
Interactive demo paper:
botcs.github.io/reason-to-pl…
3
26
141
15,803
Botos Csabi retweeted
May 11
I'm excited to share my new work joint with @csaba_botos: how well do frontier LRMs like DeepSeek-V4 learn novel video games and predict brain signatures associated with game learning? We found that they achieve SOTA on both game performance and brain predictivity! Thread 1/7
May 11
Jeee 🐦⬛
I am very proud of our joint effort with @sreejan_kumar Kumar on the project "Reason to Play"
LRMs show human-like rule discovery, and their hidden states predict human brain activity during gameplay.
Interactive demo paper:
botcs.github.io/reason-to-pl…
1
6
29
2,540
May 11
Jeee 🐦⬛
I am very proud of our joint effort with @sreejan_kumar Kumar on the project "Reason to Play"
LRMs show human-like rule discovery, and their hidden states predict human brain activity during gameplay.
Interactive demo paper:
botcs.github.io/reason-to-pl…
3
26
141
15,803
May 11
My rockstar team:
@sreejan_kumar, Austin Andrews, @LHuntNeuro, Chris Summerfield, Josh Tenenbaum, @somnirons, @marcelomattar and the brilliant @momchiltomov
The project was funded by: encode.pillar.vc/
6
427
Mar 22
Isn't harness building, context editing and prompt search analogous to LoRA, but in a semantically meaningful subspace?
It's just cheaper (no need to host the model), safer (catastrophic forgetting less likely), more interpretable (you can read now what lora would do!)
1
1
178
Jan 8
it's funny to see how mankind got advantage by fencing off a plot of land, taking it into his head to say ‘This is mine’ and now thousands of years later to advance science we need to remove the metaphorical fence, let messy nature back in to access more fruits of our land
103
Jan 6
shower thought: how would you go about teaching a dog to teach other dogs a trick?
99
28 Nov 2025
all the soap opera around NeurIPS, ICML and ICLR is a direct cause of real world forces acting on an artificial community. Having a paper can get you a job, a grant, a promotion, regardless whether the paper itself is pure garbage or seminal breakthrough.
6
881
19 Oct 2025
youtu.be/lXUZvyajciY?si=o2P4…
@karpathy Probably, the cerebellum *is* important, just mostly overlooked
@carobellum has an awesome atlas published listing even higher level functions: social, linguistic and spatial domains nature.com/articles/s41467-0…
1
1
132
22 Jun 2025
Neurips R2 thought: why is the main objective for building AI scientists to accelerate new discovery while the reproducibility crisis is already crippling many fields.
Maybe less ambitious but more robust and likely to be impactful work would be automating reproducibility check.
1
3
200
22 Jun 2025
the bonus bi-product would be some excellent training data for the future AI scientist agents.
Same way as we learned the dark arts of ML by just reimplementing famous papers in the first year of the PhD.
86
10 Mar 2025
oh my! real art is not running off the edge of the world but stopping where everyone else walks past
10 Mar 2025
We made wave dynamics flexible by adding learned damping and natural frequency encoders, allowing hidden state dynamics to adapt based on the input stimulus. On simple polygon images, we found the model learned to use these parameters to produce shape-specific wave dynamics:
6/14
1
13
2,442
14 Dec 2024
Amazing experience with the prettiest view @NeurIPSConf presenting Label Delay botcs.github.io/label-delay/
the original session was 3 hours - stayed 2h more for follow-up questions, people's interest in the topic was way above my expectations :) feeling awesome!
7
289
8 Dec 2024
Yay 🦜
Heading to Vancouver on Monday to present my first ever @NeurIPSConf paper.
I am super duper excited to finally join the cool kids’ gang.
Check out the tutorial I procrastinated over for 3 months:
youtu.be/EbPrlGkLxGo?si=qKYp…
1
16
443
Botos Csabi retweeted
22 Nov 2024
Call For Tasks for IOAI 2025: ML olympiad for high schoolers.
I know you all have strong opinions on the ML topics future computer scientists should engage with to build mastery. This is a high-impact opportunity to help us by proposing tasks.
ioai-official.org/call-for-t…
4
6
1,776
Botos Csabi retweeted
15 Aug 2024
Teams of Hungary took 5th and 6th place in the first International Olympiad in AI taking the top 2 spots for silver 🥈.
Congrats to the participants who brought a lot of self-taught expertise, adaptibility to the event continuing to build our tradition of problem solving. 🇭🇺🇭🇺🇭🇺
6
4
36
4,218
25 May 2024
What's the Magic Word? A Control Theory of LLM Prompting
arxiv.org/abs/2310.04444
Solid take by Bhargava et al, showcased on @MLStreetTalk
Application of this on AlphaFold2 could be insane - let's say you want to design a new antibody with a RNA sequence that's not patented.
2
8
1,303
23 May 2024
At HK this week - the Uni campus is jaw dropping, and when it rains it actually feels good. Is this even legal?
1
193
Botos Csabi retweeted
17 May 2024
Gen AI is poised to transform many fields, sparking major debates over its risks & calls for tighter regulation.
❗Over-regulation could be catastrophic to open-source Gen AI.
🚀 Our paper (arxiv.org/pdf/2405.08597) argues the benefits of open-source Gen AI outweigh its risks. 🧵
4
33
111
21,229
Botos Csabi retweeted
9 Apr 2024
1/n 🧵Happy to be presenting our recent work on Semi-Private Learning at @satml_conf in #Toronto @UofT with @yaxi_hu @AmartyaSanyal @FannyYangETH. Work done at @ETH_AI_Center @MPI_IS
📜Paper: openreview.net/forum?id=Ps1I…
🧑💻Code: github.com/FrancescoPinto/PI…
1
9
22
3,706