iiinova — where your digital self keeps evolving

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 27/05/2026

Speaker spotlight #5: Ian Waudby-Smith, UC Berkeley. Can we perform sequential hypothesis testing while adaptively choosing which arm to sample? Ian will discuss recent work on multi-armed sequential testing by betting.

100

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 25/05/2026

Speaker spotlight #5: Koulik Khamaru, Rutgers University. Can regularization make contextual bandits more stable? Koulik will present Stabilizing Bandit Algorithms via Regularization, on adaptive algorithms that combine low regret with reliable statistical inference.

100

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 23/05/2026

Speaker spotlight #4: Gergely Neu, ICREA & UPF. Can stochastic control offer a new route to generative modeling? Gergely will present Value-Driven Transport, a stochastic-control approach to generative modeling via efficient sample transport.

101

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 15/05/2026

Speaker spotlight #3: Kelly Zhang, Imperial College London. Can we build valid confidence intervals for the average reward of adaptive bandit algorithms? Kelly will present Bandit Simulation for Inference, a framework for inference with on-policy or off-policy bandit data.

100

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 11/05/2026

Speaker spotlight #2: Aurélien Bibaut, @netflix.com. Can policy-based contextual bandits achieve fast regret without realizability? Aurélien will explore this at AAE through variance-aware pessimistic policy learning and self-normalized martingale tools for adaptive data.

100

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 09/05/2026

Speaker spotlight #1: Tor Lattimore, DeepMind. What can continuous-time diffusions teach us about RL, ML, and adaptive experimentation? Tor will explore this at AAE through sequential design and policy gradients for stochastic bandits.

100

aaeworkshop.bsky.social @aaeworkshop.bsky.social · 23/04/2026

🚀 Excited to announce the Advances in Adaptive Experimentation (AAE) Workshop! 📅 June 18–19 🔗 t.co/Zar0qWhgA8 Join us at @gatsbyucl.bsky.social in London for two days at the frontier of causal inference 🧩, adaptive experiments 🔁, and bandits/RL 🤖!

071