Anton Tsitsulin
@tsitsulin_
Followers
3K
Following
1K
Media
217
Statuses
1K
datamodels for Gemini and Gemma 🧑🍳 Research Scientist @GoogleAI past life: graph machine learning
Joined August 2016
Somewhat unsurprisingly, the Rockefeller estate is very high taste
0
0
1
Frontier-loading models.
🇰🇷 LG recently launched EXAONE 4.0 32B - it scores 62 on Artificial Analysis Intelligence Index, the highest score for a 32B model yet. @LG_AI_Research's EXAONE 4.0 is released in two variants: the 32B hybrid reasoning model we’re reporting benchmarking results for here, and a
0
0
3
graphs are everywhere for those with eyes to see.
0
0
7
ChatGPT recommending my own papers for research question is the worst sycophancy I have ever seen. if that's your limit I'm so very sorry llmbro.
0
0
1
I’m finally vindicated for hoarding cool magnets.
BREAKING: President Trump threatens to impose "200% tariffs or something" on China if they do not provide the US with magnets.
0
0
0
> OpenAI isn't paying me and couldn't care less about me.> Immediately gets jumped on by sama. what does this mean chat.
if you are a power user, please send us feature requests!. (i asked in reply to this message and they were interesting, so would like more).
0
0
2
one more reorg.one more waifu.one more banana.beautiful week.
2
1
19
RT @jacobaustin132: Today we're putting out an update to the JAX TPU book, this time on GPUs. How do GPUs work, especially compared to TPUs….
0
536
0
OAI's 5 pro actually feels professional, Gemini DeepThink is a McKinsey consultant on NZT-48.
1
0
4
generational wealth to be made with gemma 3 270m*. *not financial advice.
Interesting convo last night at a house party. A low level employee at a large AI company (not saying which one) was there and the conversation moved to these insane salaries that the top guys are accepting to go to META and even more surprisingly, the guys that are turning down.
1
0
6
you build AI slopfriends.I finetune 270m Gemma to have a soul. we are not the same.
0
0
9
every frontier model attempts to model and answer the question, but Grok goes full reddit "expert" mode.
0
0
2
the whole drama is important for the EU sovereignty angle, if this is the optimal strategy, there is little future already.
0
0
0
underrated aspect of LLMs is that you can cheaply change *any* website with minimal effort, just a prompt to tampermonkey script.website updates and the extension breaks? boohoo, it's an extra query to your favorite LLM provider.
0
0
0
you will be spoonfed intelligence and you will like it.
the percentage of users using reasoning models each day is significantly increasing; for example, for free users we went from <1% to 7%, and for plus users from 7% to 24%. i expect use of reasoning to greatly increase over time, so rate limit increases are important.
0
0
4