Lightnews — Scholar-powered news

Reposted by Riccardo Zamboni

RL & Agents Reading Group @rl-agents-rg.bsky.social · 19d

📢 RL reading group TODAY @ 15:00 BST (in 2 hours!) 📢

Speakers: Olya Mastikhina and Dhruv Sreenivas (University of Montreal & Mila - Quebec AI Institute)

Title: Optimistic critics can empower small actors🦸

Details: edinburgh-rl.github.io/reading-group

2 2

Reposted by Riccardo Zamboni

RL & Agents Reading Group @rl-agents-rg.bsky.social · Sep 3

📢 RL reading group Thursday @ 16:00 BST 📢

Speaker: Alex Lewandowski

Title: The World Is Bigger: A Computationally-Embedded Perspective on the Big World Hypothesis 🌍

Details: edinburgh-rl.github.io/reading-group

UoE RL Reading Group

University of Edinburgh Reinforcement Learning Reading Group

edinburgh-rl.github.io

3 6

Riccardo Zamboni @ricczamboni.bsky.social · Aug 3

I really believe this work to be the tip of the iceberg! If you are at @rl-conference.bsky.social, please don’t mind disturbing Kale-ab as much as possible!

Kale-ab Tessera @kale-ab.bsky.social · Aug 3

🇨🇦 Heading to @rl-conference.bsky.social next week to present HyperMARL (@cocomarl-workshop.bsky.social) and Remember Markov (Finding The Frame Workshop).

If you are around, hmu, happy to chat about Multi-Agent Systems (MARL, agentic systems), open-endedness, environments, or anything related! 🎉

1 6

Riccardo Zamboni @ricczamboni.bsky.social · Jul 25

Loved to see these (5!!!) papers around, I really do hope it won’t turn out to be a cult.

Mirco Mutti @mircomutti.bsky.social · Jul 24

Walking around posters at @icmlconf.bsky.social, I was happy to see some buzz around convex RL—a topic I’ve worked on and strongly believe in.

Thought I’d share a few ICML papers on this direction. Let’s dive in👇

But first… what is convex RL?

🧵

1/n

3

Riccardo Zamboni @ricczamboni.bsky.social · Jul 23

Very nice and neat! If you happen to be at RLC @kale-ab.bsky.social will present our position paper on a slightly related topic! (Post to come, I was asked not to release any spoiler)

2

Riccardo Zamboni @ricczamboni.bsky.social · Jul 19

Do not worry about that, you will get free money!

1

Reposted by Riccardo Zamboni

Gaspard Lambrechts @gsprd.be · Jul 16

At #ICML2025, we will present a theoretical justification for the benefits of « asymmetric actor-critic » algorithms (#W1008 Wednesday at 11am).

📝 Paper: hdl.handle.net/2268/326874
💻 Blog: damien-ernst.be/2025/06/10/a...

ICML poster of the paper « A Theoretical Justification for Asymmetric Actor-Critic Algorithms » by Gaspard Lambrechts, Damien Ernst and Aditya Mahajan.

3 8

Reposted by Riccardo Zamboni

Onno Eberhard @onnoeberhard.com · Jul 16

I am in Vancouver at ICML, and tomorrow I will present our newest paper "Partially Observable Reinforcement Learning with Memory Traces". We argue that eligibility traces are more effective than sliding windows as a memory mechanism for RL in POMDPs. 🧵

3 12 59

Riccardo Zamboni @ricczamboni.bsky.social · Jul 10

The RL & Agents Reading Group is back with a new style 💇‍♂️! Really recommend joining the mailing list to get to know the great upcoming talks like Mattie’s

RL & Agents Reading Group @rl-agents-rg.bsky.social · Jul 10

We are super excited to kick things off again with Mattie Fellows (Postdoc @ FLAIR in Oxford) today 15:00 BST!

Paper: Simplifying Deep Temporal Difference Learning

Check out our website for full info edinburgh-rl.github.io/reading-grou...

2

Riccardo Zamboni @ricczamboni.bsky.social · Jul 8

(N/N) Come find us!
🗓️Wed. 16 July 5.30-8.30
📍B2-B3 #W-601
🔗 arxiv.org/abs/2505.01336

Enhancing Diversity in Parallel Agents: A Maximum State Entropy Exploration Story

Parallel data collection has redefined Reinforcement Learning (RL), unlocking unprecedented efficiency and powering breakthroughs in large-scale real-world applications. In this paradigm, $N$ identica...

arxiv.org

Riccardo Zamboni @ricczamboni.bsky.social · Jul 8

(4/N) Then, we also show how this diversification enforces experience diversity and better performances in downstream tasks!

1

Riccardo Zamboni @ricczamboni.bsky.social · Jul 8

(3/N) First, we show *how* to properly enforce behavioral diversity with parallel agents throught the lenses of Maximum State Entropy. The trick? Optimize over the *right* state distribution🎛️!

1 2

Riccardo Zamboni @ricczamboni.bsky.social · Jul 8

(2/N) If you can collect data on many parallel environments, would you rather use single common policy or a set of diverse, specialized ones?

1

Riccardo Zamboni @ricczamboni.bsky.social · Jul 8

🌟🌟Good news for the explorers🗺️!
Next week we will present our paper “Enhancing Diversity in Parallel Agents: A Maximum Exploration Story” with V. De Paola, @mircomutti.bsky.social and M. Restelli at @icmlconf.bsky.social!
(1/N)

1 1 4

Reposted by Riccardo Zamboni

Kale-ab Tessera @kale-ab.bsky.social · May 24

For many, the Indaba was their first exposure to ML/AI, and this helps give more people that opportunity. Please donate if you can: gofund.me/61b012e4

Deep Learning Indaba @deeplearningindaba.bsky.social · May 23

🪘Our Crowdfunding Campaign is LIVE! 🪘
🚀Support Indaba 2025 Participant Grants! 👉https://gofund.me/61b012e4

We are launching a crowdfunding campaign to directly fund essential travel and participation grants for students and researchers for the upcoming Indaba in Kigali, Rwanda.

Deep Learning Indaba: Dakar, Senegal 2024 Highlight Reel

YouTube video by Deep Learning Indaba

youtu.be

1 4

Riccardo Zamboni @ricczamboni.bsky.social · Apr 16

🚨🚨🚨
At the RL group (rl.airlab.deib.polimi.it) in Milan, Alberto M. Metelli is hiring! Postdocs and PhD students to work on the theory and practice of human feedback in RL.
When: starting this Fall.
Deadline: May 31st.
A great person to work with, my word. Spread the word!

Jobs Opportunities

albertometelli.github.io

2

Reposted by Riccardo Zamboni

EWRL18 @ewrl18.bsky.social · Apr 14

Working on something cool in RL? Submit it to EWRL!
We’ll be opening submissions soon — and this year we’re introducing a fast track for papers already accepted at other conferences! 🚀
Check out the full Call for Papers on the website: euro-workshop-on-reinforcement-learning.github.io/ewrl18/

EWRL18

euro-workshop-on-reinforcement-learning.github.io

1 5 14

Reposted by Riccardo Zamboni

Tim van Erven @timvanerven.nl · Mar 27

⏰⏰Theory of Interpretable AI Seminar ⏰⏰

In two weeks, April 8, Mirco Mutti will talk about "A Classification View on Meta Learning Bandits"

3 6

Reposted by Riccardo Zamboni

Claire Vernade @claireve.bsky.social · Jan 30

The call for workshops is out. Now may be a bit busy but the deadline is on March 7th so there’s time :)
Send us your great ideas for 1-day workshops at RLC!

Reinforcement Learning Conference @rl-conference.bsky.social · Jan 29

RLC call for workshops is out rl-conference.cc/callforworks...!
Submissions open on Feb 3rd, deadline on March 7, with the workshops on Aug. 5th. Last year's workshops were inimitable and we @claireve.bsky.social and Josiah Hanna) look forward to your amazing proposals

RLJ | RLC Call for Workshops

rl-conference.cc

2 8

Reposted by Riccardo Zamboni

Gergely Neu @neu-rips.bsky.social · Jan 20

my 4th phd student @nnekaokolo4.bsky.social just graduated today. i have no words for how proud i am of her and the work we've done together

3 3 52

Reposted by Riccardo Zamboni

Kale-ab Tessera @kale-ab.bsky.social · Jan 10

We recently asked the authors to do a talk on this - m.youtube.com/watch?v=Bmdn... . A lot of awesome gems (goal focused rl research, don't waste your time, etc) 🔥 Thanks for bringing this to my attention @brandonrohrer.com !

2 6 21

Reposted by Riccardo Zamboni

Lukas Schäfer @lukaschaefer.bsky.social · Dec 9

📚🧵1/7 It is finally here!! Only one more week until the print release of our textbook “Multi-Agent Reinforcement Learning: Foundations and Modern Approaches” with @mitpress.bsky.social!

What you get, why you should be interested and more, all below in a short 🧵👇

4 19 66

Riccardo Zamboni @ricczamboni.bsky.social · Nov 27

Hi! I’d be happy to join these cool folks :)

1 1

Riccardo Zamboni @ricczamboni.bsky.social · Nov 27

Would be happy to join :)

1 1