Núria Armengol
@NriaArmengol2
Followers
134
Following
243
Media
5
Statuses
66
ETH/CLS PhD candidate focused on reinforcement learning and sports lover.
Joined February 2017
We just had so much fun to replay human motions on a robot. 👯♀️
8
40
235
🚀We propose a method for active exploration in unsupervised RL, guiding data collection towards minimizing the uncertainty that the algorithm has of its representations. Joint work with @vlastelicap , @GMartius and Stelian Coros Still interested? Paper
0
0
2
❓In our work we wonder: "What is the best data collection strategy to increase sample efficiency when learning a generalist representation of all optimal policies?"
1
0
1
🌐The goal of zero-shot RL is to provide algorithms for recovering optimal policies for all possible reward functions given interaction data with the environment.
1
0
1
Last week I presented our last work: 🐝“Epistemically-guided forward backward exploration (FBEE)”🐝 at the @RL_Conference TLDR: Active learning for unsupervised RL
2
9
50
When multiple tasks need improvements, fine-tuning a generalist policy becomes tricky. How do we allocate a demonstration budget across a set of tasks of varied difficulty and familiarity? We are presenting a possible solution at ICML on Wednesday! (1/3)
1
8
17
✨Introducing SENSEI✨ We bring semantically meaningful exploration to model-based RL using VLMs. With intrinsic rewards for novel yet useful behaviors, SENSEI showcases strong exploration in MiniHack, Pokémon Red & Robodesk. Accepted at ICML 2025🎉 Joint work with @cgumbsch 🧵
2
36
152
Consider submitting your talk proposals to EWRL! 🐣 More info in 🦋 @ewrl18.bsky.social and the website
euro-workshop-on-reinforcement-learning.github.io
EWRL News!⚡ We’re pleased to announce a new Call for Contributed Talks to bring novel and diverse viewpoints to EWRL2025. Early-career researchers are especially encouraged to submit proposals and share their work with the community. Full details: https://t.co/ClEKyTz4zb
0
3
7
EWRL is happening in Tübingen this year! Checkout our 🦋 posts. When? September 17-19, 2025. https://t.co/I73FwkO9WM
0
4
6
I am very proud and excited about this project! We are learning a general particle simulator directly from videos end-to-end. Check it out 👇 #Simulation #Gausiansplatting #ML
Introducing 3DGSim🧩— an end-to-end 3D physics simulator trained only on multi-view videos. It achieves spatial & temporal consistency w/o ground truth 3D info or heavy inductive biases— enabling scalability & generalization🚀 🔗 https://t.co/OTJGrxZqyk 📺 https://t.co/4ORWgnO3qv
0
1
14
🗣️ "Metges i empreses ens venien a buscar per fer servir la tecnologia, l'única manera per transferir-la era crear una spin-off". Descobreix l'entrevista amb l'#alumniUPF @JordiMill, cofundador de #VirTest, i investigador i emprenedor d'@enginyeria_upf. 👉
0
4
5
We are excited to share that Fatemeh Zargarbashi will present her recent paper "Robot Keyframing: Learning Locomotion with High-Level Objectives via Mixture of Dense and Sparse Rewards," at the upcoming CoRL 2024 @corl_conf.
3
27
139
Last Friday, the Prorok Lab, led by Prof. Amanda Prorok @aprorok from the University of Cambridge @Cambridge_Uni, visited CRL @crl_ethz for a collaborative workshop. Both groups engaged in insightful discussions, fostering an enriching exchange of ideas.
0
3
10
I was in a 1:1 meeting with a colleague and he shared this article with me: “It takes two to think”. I wish I reminded myself of this more often. To all phd students out there, it’s worth a read!
nature.com
Nature Biotechnology - It takes two to think
0
0
9
Become a Center For Learning Systems doctoral researcher! The application portal is now open until Nov 5th: https://t.co/pqfeB7uj83 It's a fully-funded doctoral fellowship program with co-supervision by one advisor from ETH Zürich ... (1/2)
3
14
63
📣Ph.D. Fellowship opportunities at the @ETH_en @MPI_IS Center for Learning Systems (CLS) joint doctoral program! Deadline November 5 2024
learning-systems.org
5
58
243
GSO 2025 in Tübingen, Germany -- 12-14. Feb. *Guided Self-organization -- Machine Learning in Embodied Agents.* Deadline 27. September, submit an extended abstract! Looking forward to great discussions. #MachineLearning #Complexsystems #Robotics #InformationTheory #Conference
2
8
22
Were you ever wondering why there are little-to-no diverse offline imitation learning algorithms 🤔? Then we've got something 4 you! Our paper, "Offline Diversity Maximization under Imitation Constraints" is being presented today at RLC2024 🎇! 🧵 https://t.co/NTOedhW9vD
1
18
76
🧵 8/9 Pass by our poster #1100 at ICML today Thursday at 11:30h. Authors 🤹: Núria Armengol, Marco Bagatella (@mar_baga ), Marin Vlastelica (@vlastelicap) and Georg Martius (@GMartius). Paper 📄 : https://t.co/PR2If9C85t Project website 🕸:
1
0
2