Aldo Pacchiano
@aldopacchiano
AI research at Broad Institute and Boston University. Reinforcement Learning / Bandits / Experiment Design Mexicano 🇲🇽
An overview of Online Model Selection results for contextual bandits and RL. Presented at the UCL Statistical Science Seminar. youtube.com/watch?v=R8xMTG…
Going to #ICML2025 next week, ping me up if you are interested in pure exploration problems in #RL! Also check my poster on Exploration for Policy Evaluation, Tuesday morning 11.30am! Joint work with @aldopacchiano at BU.
Our new work on how to Learn-to-Explore using Transformer Models will be presented today the EXAIT workshop at #ICML2025! This is joint work with Alessio Russo @rssalessio and Ryan Welch.
Wondering how to do online pure exploration using transformer models? Check our EXAIT Workshop at #ICML2025, presented today by my brilliant co-author Ryan Welch Full paper: arxiv.org/abs/2506.01876 A joint work with @aldopacchiano @ BU/MIT/Broad institute.
Had an amazing session at #ICML2025! Happy to have presented our last paper on exploration for policy evaluation in #RL, a joint work with @aldopacchiano at #BU Check the paper here arxiv.org/abs/2502.02516 And follow us for more work on principled exploration in RL!
AGI is going to be completed by online learning (active learning & RL from experience), not offline learning. Even pre-training is best informed by downstream product feedback.
AGI is going to be achieved by a product, not necessarily a “model”.
Excited to share this work on efficient policy evaluation with @aldopacchiano got accepted at #ICML2025. Adaptive Exploration for Multi-Reward Multi-Policy Evaluation arxiv.org/pdf/2502.02516 Check github repo here github.com/rssalessio/mul… #ML #RL #ICML
Our work "On the Hardness of Bandit Learning" co-authored with Nataly Brukhim, Miro Dudik @MiroDudik and Robert Schapire was accepted into COLT 2025!
Our work "Pure Exploration with Feedback Graphs" is being presented by Alessio @rssalessio at AISTATS 2025!
A witty student drew this in the final exam of the bandit class I’m teaching.
Our new LLM personalization work is out :) ... led by amazing collaborators @IdanShenfeld and Felix Faltings and joint with @pulkitology .
The next frontier for AI shouldn’t just be generally helpful. It should be helpful for you! Our new paper shows how to personalize LLMs — efficiently, scalably, and without retraining. Meet PReF (arxiv.org/abs/2503.06358) 1\n
My group @FLAIR_Ox is recruiting a postdoc and looking for someone who can get started by the end of April. Deadline to apply is in one week (!), 19th of March at noon, so please help spread the word: my.corehr.com/pls/uoxrecruit…
New paper “Pure Exploration with Feedback Graphs” with BU postdoc Alessio Russo @rssalessio and BU PhD student Yichen Song. Accepted for oral presentation @ AISTATS 2025.
Thrilled to share this work with Yichen Song and @aldopacchiano got accepted for an oral presentation at AISTATS 2025. Pure Exploration with Feedback Graphs arxiv.org/pdf/2503.07824 Code: github.com/rssalessio/Pur… #AISTATS #AISTATS2025 #ML
Casting reward selection as a model selection leads up to 8x faster learning and 50% better performance! (arxiv.org/abs/2410.13837) ⚡ Provable regret guarantees. 🌟 Easy to implement (github.com/Improbable-AI/…). ⚔️ 1 GPU can do the work of up to 8 GPUs! Presenting ORSO:…
Excited to share our new survey of in-context reinforcement learning!! arxiv.org/abs/2502.07978 w/ @AmirMoeini99 @wangjiuqi @jakeABeck @EthanBlaser @shimon8282 @rohanchandra30