Self-supervised Reinforcement Learning

concept
2 mentions 1 recordings first heard Jan 2026 last heard 2 Jan ↑2 vs the 12 months before

Approach that learns representations without explicit rewards

Trend

mentions per week · public audio 7d30d6M12M
2 · 29 Dec SepDecMarJunnow

Peak in the week of 29 Dec — 2 mentions across 1 series.

Moments

newest first · ▶ plays the moment
Kevin WangLatent Space: The AI Engineer Podcast · [NeurIPS Best Paper] 1000 Layer Networks for Self-Supervised... · 5:07 · 2 Jan
…So can self-supervised RL scale in similar ways?…
· Open transcript →
Kevin WangLatent Space: The AI Engineer Podcast · [NeurIPS Best Paper] 1000 Layer Networks for Self-Supervised... · 4:37 · 2 Jan
…So we tried a different approach to RL, um, called self-supervised RL, where instead of learning like a value function, we're learning representations of states, actions, and future states, such that the representations along the same trajectory are pushed together, the representations.…
· Open transcript →
All 2 moments