![Brendan O'Donoghue Profile](https://pbs.twimg.com/profile_images/1379478159402397697/IImTOUYq_x96.jpg)
Brendan O'Donoghue
@bodonoghue85
Followers
3K
Following
77K
Statuses
2K
Research scientist at @GoogleDeepMind, working on generative models, deep learning, RL. PhD from @stanford.
London
Joined August 2010
RT @JeffDean: Delighted to be a minor co-author on this work, led by @pranavn1008: Combining losses for different Matyroshka-nested groups…
0
34
0
RT @euaccofficial: Top LLM models sources of funding: - OpenAI, private investment - Claude, private investment - DeepSeek, private investm…
0
482
0
RT @JeffDean: Training our most capable Gemini models relies heavily on our JAX software stack + Google's TPU hardware platforms. If you…
0
161
0
@giffmana @EU_Commission It has an award Lucas! For excellence! You shouldn't need to see benchmarks!
0
0
12
RT @lmarena_ai: Breaking News: DeepSeek-R1 surges to the top-3 in Arena🐳! Now ranked #3 Overall, matching the top reasoning model, o1, whi…
0
386
0
RT @demishassabis: Project Astra is our vision for a universal AI assistant. Soon some of its capabilities will come to Android & Samsung G…
0
125
0
RT @demishassabis: Our latest update to our Gemini 2.0 Flash Thinking model (available here: scores 73.3% on AIME…
0
363
0
RT @bodonoghue85: @conormdurkan @Dilip_Arumugam I think a lot of people get confused about this, the literature is very muddled and overloa…
0
2
0
I think a lot of people get confused about this, the literature is very muddled and overloads terminology a lot. Take for instance the recent RL overview by @sirbayes . Here "Posterior sampling RL", a real Bayesian inference procedure, is presented *immediately before* "RL as a posterior inference problem", a framework that has nothing to do with Bayesian RL. A reader would be forgiven for thinking they were related, (the words are the same just rearranged!). In fact they are not at all related and the second one is not a principled approach to handling uncertainty in RL, despite the name. Myself and @IanOsband have been banging this drum for a while now with little success. I have spoken with Kevin about this and I am hoping he updates the monograph asap 🙏🙏🙏
0
2
11
RT @matthewclifford: 1/ Today the Prime Minister published my AI Opportunities Action Plan for the UK and committed to implement all its re…
0
257
0
RT @ziv_ravid: 🧵 Just read the fascinating Microsoft Research paper on rStar-Math. They achieved OpenAI o1-level math performance with tiny…
0
65
0
RT @stanley_h_chan: Tutorial on Diffusion Models for Imaging and Vision has been published. Official copy: Free a…
0
140
0
RT @douwekiela: I’m really sad that my dear friend @FelixHill84 is no longer with us. He had many friends and colleagues all over the world…
0
92
0
RT @jk_rowling: The details emerging about what the rape gangs (why call them 'grooming' gangs? It's like calling those who stab people to…
0
26K
0
RT @demishassabis: NotebookLM is an awesome tool for summarising and researching complex topics. I use it all the time for my research, it'…
0
329
0
RT @jachiam0: If you want to know something deep, fundamental, and maximally portable between virtually every field: study mathematical opt…
0
148
0
Do NOT tell Vivek Ramaswamy about this
10 yo neighbor kid was out in 48 degree weather and heavy rain with his new Duck helmet and gloves he got for Christmas. I stood on the porch having a smoke, watching him run routes and being qb and receiver by himself as the rain came down. I'm 58 and said screw it. I grabbed my grandsons Nerf, and called him over. Wanna run some routes kid? Him... toss it. What a great Christmas. We got wet. I came in and my wife just smiled. No grief for getting the floor wet. Good kid.
1
0
2
RT @kimmonismus: Veo 2: "A lego shark swims past a real coral reef while it chases a real scuba diver. nature photography, 35mm film." via…
0
108
0