Yujia Li Profile
Yujia Li

@liyuajia

Followers
3K
Following
595
Statuses
118

Machine Learning

Joined August 2011
Don't wanna be here? Send us removal request.
@liyuajia
Yujia Li
13 days
The translation is terse compared to the original Chinese version which has a bit more casual tone.
0
0
0
@liyuajia
Yujia Li
3 months
0
0
0
@liyuajia
Yujia Li
4 months
@kcjacksonwang Congrats! And amazing use case šŸ˜
1
0
1
@liyuajia
Yujia Li
5 months
RT @RemiLeblond: Cool results from OpenAI: 89th percentile on Codeforces! A year ago #AlphaCode2 was at the 85th. Goes to show how hard thiā€¦
0
4
0
@liyuajia
Yujia Li
1 year
RT @JeffDean: Gemini 1.5 Pro - A highly capable multimodal model with a 10M token context length Today we are releasing the first demonstrā€¦
0
1K
0
@liyuajia
Yujia Li
1 year
RT @jacobaustin132: We've finally put out a detailed IEEE/ACM paper on @Google's multi-year effort to ease the burden of code review with Mā€¦
0
142
0
@liyuajia
Yujia Li
1 year
@TacoCohen Congrats!
0
0
1
@liyuajia
Yujia Li
1 year
A great example of using LLMs to search effectively in the space of programs describing algorithms, to discover new scientific knowledge!
@GoogleDeepMind
Google DeepMind
1 year
Introducing FunSearch in @Nature: a method using large language models to search for new solutions in mathematics & computer science. šŸ” It pairs the creativity of an LLM with an automated evaluator to guard against hallucinations and incorrect ideas. šŸ§µ
0
1
7
@liyuajia
Yujia Li
1 year
RT @sundarpichai: Today developers can start building with our first version of Gemini Pro through Google AI Studio at
0
1K
0
@liyuajia
Yujia Li
1 year
RT @cHHillee: As mentioned previously, I found AlphaCode2 accounts, and through stalking their submission history, I manually performed theā€¦
0
63
0
@liyuajia
Yujia Li
1 year
@KL_Div Thanks!
0
0
0
@liyuajia
Yujia Li
1 year
Also, excited to see the next iteration of AlphaCode released! Fine-tuned on top of Gemini, this new model is better than ~85% of competitive programmers, a huge improvement from the estimated 46% in the initial AlphaCode work published last year, and it can definitely beat me!
@RemiLeblond
RĆ©mi Leblond
1 year
So excited to share what the team and I have been working on these last months! #AlphaCode 2 is powered by Gemini and performs better than 85% of competition participants in 12 contests on Codeforces! More details at @GoogleDeepMind
Tweet media one
0
0
21
@liyuajia
Yujia Li
1 year
RT @PeterWBattaglia: Weā€™ve released GraphCast's code and weights, so everyone can run it:
0
92
0
@liyuajia
Yujia Li
1 year
Really excited to see #Lyria released! An amazing model and great tool that can be used to create music and a lot of fun. Looking forward to a custom radio station of generated music tailored just for mešŸ˜‰ Behind this is a lot of hard work by an amazing team, congratulations!
@demishassabis
Demis Hassabis
1 year
Thrilled to share #Lyria, the world's most sophisticated AI music generation system. From just a text prompt Lyria produces compelling music & vocals. Also: building new Music AI tools for artists to amplify creativity in partnership w/YT & music industry
0
1
18
@liyuajia
Yujia Li
2 years
Come talk to us at ICML about our latest work on positional encodings for directed graphs!
@geisler_si
Simon Geisler
2 years
The transformer architecture powers recent AI tools like #ChatGPT or #GoogleBard. In our @DeepMind #ICML2023 paper Transformers Meet Directed Graphs, we generalize transformers to more general inputs, namely directed graphs. Hereā€™s how we did it. šŸ§µ
Tweet media one
0
3
43
@liyuajia
Yujia Li
2 years
Code optimization is a great research area for code + search. In optimization, correctness can be checked using the original code. Efficiency is a clear reward signal. When optimizing important code used everywhere, small changes deliver huge impact. Glad to be part of this!
@GoogleDeepMind
Google DeepMind
2 years
As society becomes more digital, itā€™s critical to improve the code powering the world's computing. Today in @Nature, we present AlphaDev, an AI system using reinforcement learning to discover enhanced computer science algorithms. How does it work? šŸ§µ
0
1
7
@liyuajia
Yujia Li
2 years
RT @jacobaustin132: Excited to see a blog post on one of the coolest projects I've worked on at Google: using LLMs to automatically resolveā€¦
0
71
0
@liyuajia
Yujia Li
2 years
RT @geisler_si: If you are at #aaai2023 come to todayā€˜s workshop Deep Learning on Graphs (room 145B)! I will be presenting our work on tranā€¦
0
11
0