Tweetovi
- Tweetovi, trenutna stranica.
- Tweetovi i odgovori
- Medijski sadržaj
Blokirali ste korisnika/cu @janleike
Jeste li sigurni da želite vidjeti te tweetove? Time nećete deblokirati korisnika/cu @janleike
-
Prikvačeni tweet
The agent alignment problem may be one of the biggest obstacles for using ML to improve people’s lives. Today I’m very excited to share a research direction for how we’ll aim to solve alignment at
@DeepMindAI. Blog post: https://medium.com/@deepmindsafetyresearch/scalable-agent-alignment-via-reward-modeling-bf4ab06dfd84 … Paper: https://arxiv.org/pdf/1811.07871.pdf …pic.twitter.com/33HQTlkcQR
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Lucas Perry from FLI interviewed me for their podcast! If you want to find out why I don't do theory research anymore, what's going on in safety at DeepMind, and how we're planning to solve the alignment problem, check it out: https://futureoflife.org/2019/12/16/ai-alignment-podcast-on-deepmind-ai-safety-and-recursive-reward-modeling-with-jan-leike/ …
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Rob has a whole YouTube channel dedicated to explaining AI safety ideas for everyone to understand. Definitely worth checking out if you're interested in this stuff!https://m.youtube.com/channel/UCLB7AzTwc6VFZrBsO2ucBMg …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Very accessible explanation of the motivation behind reward modeling by
@robertskmiles:https://youtu.be/PYylPRX6z4QPrikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Below you can see our algorithm's generated trajectories in
@OpenAI's Car Racing task. They don't need to be 100% realistic, just good enough for the human to label them correctly (e.g. driving into the grass is bad).pic.twitter.com/KxSd3opS7aPrikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
How do you train an RL agent in the presence of unknown, unsafe states without visiting them even once? New algorithm by our intern
@sidgreddy synthesizes trajectories with a generative model and ask a human to label them for safety.https://deepmind.com/blog/article/learning-human-objectives-by-evaluating-hypothetical-behaviours …Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
Want to ensure AI is beneficial for society? Come talk to like-minded people at the Human-Aligned AI Social at
#NeurIPS2019, Thursday 7-10 pm, room West 205-207. https://nips.cc/Conferences/2019/Schedule?showEvent=15974 …@claudia_shi57@victorveitchpic.twitter.com/0KgrHGZSiu
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
I love all of the programmes, but I particularly like ep 4 - one of my favourite bits of the whole series is when
@janleike explains the human preferences work we did with@OpenAIhttps://twitter.com/DeepMindAI/status/1163829592265805824 …
1:55Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
How do you design agents that don’t have an incentive to tamper with their reward signal?
@tom4everitt et al. derive design principles for RL algorithms. Easy fix if you’re doing model-based RL!https://medium.com/@deepmindsafetyresearch/designing-agent-incentives-to-avoid-reward-tampering-4380c1bb6cd …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
Excited to finally share this AI Reading List, compiling key resources on artificial intelligence & its long-term implications. The list is divided into "80/20" resources and "deep dive" resources to help with suggested prioritization.https://medium.com/@v_maini/ai-reading-list-c4753afd97a …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
It's 9:15am in room 104. https://icml.cc/Conferences/2019/Schedule?showEvent=4339 …
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Very excited to deliver the
#icml2019 tutorial on#safeml tomorrow together with@csilviavr! Be prepared for fairness, human-in-the-loop RL, and a general overview of the field. And lots of memes!pic.twitter.com/o0zrDPLYxo
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
I had a fantastic conversation with
@DeepMindAI scientist Pushmeet Kohli about how they're developing new ways to keep AI systems robust & reliable, why it’s a core issue in AI design that everyone has to attend to, and how to succeed as an AI researcher:https://80000hours.org/podcast/episodes/pushmeet-kohli-deepmind-safety-research/ …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
How do we uncover failures in ML models that occur too rarely during testing? How do we prove their absence? Very excited about the work by
@DeepMindAI’s Robust & Verified AI team that sheds light on these questions! Check out their blog post:http://deepmind.com/blog/robust-and-verified-ai …Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
New version on arXiv of our ICLR paper with
@DBahdanau,@FelixHill84,@janleike, Edward Hughes, and@pushmeet. We jointly learn language-conditional policies and reward models. Updated results/explanations + discussion of relation with other IRL methods. https://arxiv.org/abs/1806.01946 pic.twitter.com/BXkLrGY6UC
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Recent evidence that there can be unexpected unaligned agents in your data center: https://twitter.com/mtellin/status/1088995712736600065 …
Tweet je nedostupan.Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
Join us and
@Blizzard_Ent this Thursday at 6:00pm GMT for an exciting#StarCraft demonstration, hosted by@Artosis and@RotterdaM08! Livestream on YouTube: https://www.youtube.com/c/deepmind Read more about#StarCraft2 as an environment for AI research: https://deepmind.com/blog/deepmind-and-blizzard-open-starcraft-ii-ai-research-environment/ …pic.twitter.com/Eztc5Bro5Y
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Multiparty computation is awesome because it lets multiple parties train a model without seeing the weights. But there are fundamental limits to making it scalable: >24x overhead! Our new paper addresses this problem. https://arxiv.org/abs/1812.05979 w/
@MiljanMartic@iamtrask et al.pic.twitter.com/xqkujqKUdx
Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Jan Leike proslijedio/la je Tweet
Good morning
#NeurIPS18! Stop by our recruitment stand from now until 9:50am to meet with our Safety team. Read more about their work: https://medium.com/@deepmindsafetyresearch/building-safe-artificial-intelligence-52f5f75058f1 … Later today, our Science team will have a meet & greet at 3pm - come chat about protein folding!https://deepmind.com/blog/alphafold/Hvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi -
Finally, more evidence that the reward model needs to be trained with humans in the loop; otherwise the agent learns to exploit the reward model, for example by pretending to shoot a spider in Hero. 3/3pic.twitter.com/AYRBBtIrJh
Prikaži ovu nitHvala. Twitter će to iskoristiti za poboljšanje vaše vremenske crte. PoništiPoništi
Čini se da učitavanje traje već neko vrijeme.
Twitter je možda preopterećen ili ima kratkotrajnih poteškoća u radu. Pokušajte ponovno ili potražite dodatne informacije u odjeljku Status Twittera.

for a preview or subscribe