Dylan R. Ashley
@oneDylanAshley
PhD student studying reinforcement learning with @SchmidhuberAI. MSc with Rich Sutton of @rlai_lab. Sometimes amateur photographer. Opinions are my own.
So proud of my MSc supervisor for winning the #TuringAward (the #NobelPrize in CS). If there’s one thing that’s always set @RichardSSutton a peg above the average researcher I’ve met, it’s undoubtedly the absurd degree of his dedication to advancing human knowledge.
Definitely the longest paper I've been a part of so far. Some pretty intense theory that gives one of the first peeks into the depths of these algorithms.
Do you like RL and math? Our collaboration, IDSIA-KAUST-NNAISENSE, has the most detailed exploration of the convergence and stability of modern RL frameworks like Upside-Down RL, Online Decision Transformers, and Goal-Conditioned Supervised Learning arxiv.org/abs/2502.05672
Can confirm that it’s pretty swish. There’s some intense people that win this.
Are you a rising star in AI? 🌟 Join us as a speaker for the 4th edition of the KAUST Rising Stars in AI Symposium. In the past 2 years co-organizing this event, I've met incredible researchers now in top industrial and academic positions worldwide. More info: 📅 Event date:…
We’ll shortly be presenting at ISMIR a small follow-up to our narrative essence work, where we looked at how well a transformer can accomplish automatic album sequencing. The answer is not as well as a narrative essence approach. You can read more here: arxiv.org/abs/2411.07772
Our poster session starts in a few minutes. Come by and check out our work!
Some very cool #AI #DeepLearning #RL work to have been involved with. Interestingly, there's no clear bound on the depth of the network or general scalability here, so a lot of potential.
Some quite groundbreaking #LLM #deeplearning #AI work to have been a part of. All good frontier benchmarks should include an a non-trivial automatic evaluator.
🔔 new 𝗔𝗴𝗲𝗻𝘁-𝗮𝘀-𝗮-𝗝𝘂𝗱𝗴𝗲 paper: 𝗖𝗮𝗻 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 𝗲𝘃𝗮𝗹𝘂𝗮𝘁𝗲 𝗔𝗜 𝗮𝗴𝗲𝗻𝘁𝘀 𝗮𝘀 𝗲𝗳𝗳𝗲𝗰𝘁𝗶𝘃𝗲𝗹𝘆 𝗮𝘀 𝗵𝘂𝗺𝗮𝗻𝘀? 𝗬𝗲𝘀, 𝘁𝗵𝗲𝘆 𝗰𝗮𝗻! 📄 arxiv.org/abs/2410.10934… 👨💻 github.com/metauto-ai/age… Introducing 𝗔𝗴𝗲𝗻𝘁-𝗮𝘀-𝗮-𝗝𝘂𝗱𝗴𝗲, a…
Why pay for Claude, when I can get my code written by amazon.
Some very cool #AI #DeepLearning #RL work to have been involved with. Interestingly, there's no clear bound on the depth of the network or general scalability here, so a lot of potential.
Can neural networks with 5000 layers improve long-term planning? 🤖 Check out our latest research with @SchmidhuberAI, @oneDylanAshley, and team: arxiv.org/abs/2406.08404 #AI #DeepLearning #RL
The old city of #jeddah: #albalad instagram.com/p/CpsWgLfDXdk/…
Finally had some time to write down some thoughts. All You Need Is Supervised Learning: From Imitation Learning to Meta-RL With Upside Down RL w/ @rupspace @oneDylanAshley @SchmidhuberAI arxiv.org/abs/2202.11960
Decidated to everyone else who's had to live through the horror 💀
First, AlphaFold validates the field of Artificial Intelligence again, then nature.com/articles/s4158… validates the subfield of Reinforcement Learning. Last week was pretty good.
