Aly Lidayan
@aliday.bsky.social
AI PhD student at Berkeley
alyd.github.io
alyd.github.io
I'm presenting this 3-5:30pm on Saturday, Hall 3 #396 🌞
come chat about designing rewards and intrinsic motivation for RL + meta RL!
come chat about designing rewards and intrinsic motivation for RL + meta RL!
🚨Our new #ICLR2025 paper presents a unified framework for intrinsic motivation and reward shaping: they signal the value of the RL agent’s state🤖=external state🌎+past experience🧠. Rewards based on potentials over the learning agent’s state provably avoid reward hacking!🧵
April 25, 2025 at 2:58 AM
I'm presenting this 3-5:30pm on Saturday, Hall 3 #396 🌞
come chat about designing rewards and intrinsic motivation for RL + meta RL!
come chat about designing rewards and intrinsic motivation for RL + meta RL!
Reposted by Aly Lidayan
1/3 Out now: new paper on people's perception of AI (robot) creativity! Core finding: we attribute more creativity to a creative act if people not only see the final artwork, but also its creation process & the robot making it. Video: vimeo.com/1073134853 Open-access paper: doi.org/10.1145/3711...
April 8, 2025 at 1:27 PM
1/3 Out now: new paper on people's perception of AI (robot) creativity! Core finding: we attribute more creativity to a creative act if people not only see the final artwork, but also its creation process & the robot making it. Video: vimeo.com/1073134853 Open-access paper: doi.org/10.1145/3711...
🚨Our new #ICLR2025 paper presents a unified framework for intrinsic motivation and reward shaping: they signal the value of the RL agent’s state🤖=external state🌎+past experience🧠. Rewards based on potentials over the learning agent’s state provably avoid reward hacking!🧵
March 26, 2025 at 12:05 AM
🚨Our new #ICLR2025 paper presents a unified framework for intrinsic motivation and reward shaping: they signal the value of the RL agent’s state🤖=external state🌎+past experience🧠. Rewards based on potentials over the learning agent’s state provably avoid reward hacking!🧵