Miruna Pîslar

Senior Research Engineer @GoogleDeepMind

[Miruna Pîslar @Miruna_Pislar]

Proud to be working on the exciting LearnLM project! Check out our tech report 📖 https://t.co/KtAqAOkEzo

[Google @Google]

In a recent technical report, LearnLM, our set of AI models and capabilities fine-tuned for learning, outperformed other leading AI models on the principles of learning science. Now it’s available to try out in AI Studio. Learn more ↓ https://t.co/V6IPvlSo4n

[Katherine Hermann @khermann_]

“Towards Responsible Development of Generative AI for Education: An Evaluation-Driven Approach” is now available at https://t.co/xQygtwSXdx 

#ICML2024: Irina Jurenka and Markus Kunesch will be demoing the LearnLM-Tutor at the GDM booth on Tues afternoon: https://t.co/Hif7rD7BFU

[Miruna Pîslar @Miruna_Pislar]

Thrilled to have been part of the LearnLM journey! 🚀

[Google @Google]

The new Learning coach Gem uses LearnLM to provide step-by-step study guidance, helping you build understanding instead of just giving you an answer. It will launch in Gemini in the coming months. #GoogleIO

[Miruna Pîslar @Miruna_Pislar]

Mode-switching policies offer amazing flexibility in #ReinforcementLearning!🤖Loren Anderson's #ICLR2024 blog post provides a great analysis, building on our "When should agents explore?" paper. See what makes these policies so adaptable:
https://t.co/slIwMOS5Sh
Thank you, Loren!

[Miruna Pîslar @Miruna_Pislar]

3/3 Our method, including the AI-designed mechanism for resource allocation, has the potential to go beyond a self-designed, isolated economic game and into real-world scenarios. We believe this work can drive positive change in policy, economics, and social sciences.

[Miruna Pîslar @Miruna_Pislar]

What if AI could help us build a more cooperative society? 🌱 Our latest research reveals how a deep-RL agent incentivizes humans to behave sustainably in a common-pool resource game: https://t.co/TvdIdoSamF.
Check it out! #AI #sustainability

[Miruna Pîslar @Miruna_Pislar]

2/3 Through careful analysis, black-box neural networks can successfully be “distilled” into explainable, clear, and trustworthy mechanisms.

[summerfieldlab @summerfieldlab.bsky.social @summerfieldlab]

This project was a long time coming. We used deep RL to find distribution mechanisms that encourage humans to behave sustainably, in a common-pool resource game. Led by @Miruna_Pislar and Raphael Koster
https://t.co/YHQClSL4rZ

[Edward Hughes @edwardfhughes]

Absolutely delighted that our paper came out in Nature Communications today. This was an incredible effort from an amazingly talented team at GDM that I had the great privilege to lead. https://t.co/1OHo0H2rXG

[Miruna Pîslar @Miruna_Pislar]

Excited to share our cultural transmission research is now published in Nature Comms!
From a blank slate to cultural connoisseur, our paper reveals the journey of agents learning to navigate an open-ended world through observational learning. Check it out!
https://t.co/Xkjn4aeU1v

[Kory Mathewson @korymath]

Excited to share our @GoogleDeepMind work creating agents that can mimic experts in goal-driven behaviour in real-time 3D environments.

It's a significant stride in human-AI interaction.

Delve into the details in our paper in Nature Communications
https://t.co/HmDjUUaAtI

[Yann LeCun @ylecun]

Arthur Mensch, cofounder of French AI startup Mistral, asking @EmmanuelMacron about EU-level regulation of AI.

Answers:
1. premature regulation would stifle innovation.
2. More investment in computing infrastructure. France has an advantage: low-emission nuclear energy.
3. "We believe in open source platforms."
4. Enable the use of public data to train AI systems.

[CoML (Cognitive Machine Learning) | @ENS @CoML_ENS]

[ Seminar 🗣️] Tomorrow at 4pm PARIS TIME, we are thrilled to welcome @Miruna_Pislar from @DeepMind ! She will present us «Learning Robust Real-Time Cultural Transmission without Human Data», work studying cultural evolution as a way to achieve AGI🤖!

🖱️👉 https://t.co/vXYCEJOYIf

[Google DeepMind @GoogleDeepMind]

In a new @NatureHumBehav paper, our team trained an AI system to distribute resources according to human preferences, fostering cooperation and fairness.

Read more at https://t.co/h2iByPSb9i and https://t.co/hfoB8lrmBa 1/

[Google DeepMind @GoogleDeepMind]

Exploring in sparse-reward partially-observable tasks? BYOL-Explore harnesses the power of Self-Supervised Learning to solve such tasks only with prediction error as intrinsic reward, producing a conceptually simple yet powerful exploration agent: https://t.co/2M84ufghAP 1/

[Google DeepMind @GoogleDeepMind]

This morning @ICLRconf - an adaptive mechanism which endows an agent with diverse and useful experiences that translate to meaningful performance differences.

Learn more at 10:30AM BST: https://t.co/o4Sov5dqK9

#ICLR2022 1/

[Miruna Pîslar @Miruna_Pislar]

Excited to present our paper "When should agents explore?" @ICLRconf! Join us at Poster Session 7 to learn about the temporal aspect of exploration and how our mechanism for intra-episodic switching endows RL agents with diverse and useful experiences. https://t.co/HTod9s49la

[Google DeepMind @GoogleDeepMind]

It’s that time of year again…@ICLR_conf! Here’s a brief glimpse into our upcoming oral, spotlight, and poster presentations this week: https://t.co/psZXtOTUlx

Don’t miss our 5+ sessions kicking off this morning at 10:30AM BST. See you there! #ICLR2022

[Kory Mathewson @korymath]

New @DeepMind paper on Learning Robust Real-Time Cultural Transmission without Human Data explores the emergence of collective knowledge in reinforcement learning agents.

More details:
🌐: https://t.co/w7abrPWSob
📝: https://t.co/GNwl4wN9LB
📹: https://t.co/Oz5DD3R198