Jun 14
Title: Preserving Plasticity in Continual Learning via Dynamical Isometry
Authors: Andries Rosseau, Robert Müller (@deepqlearning), Ann Nowé
From the Deep Manifold view, dynamical isometry helps preserve plasticity, but it is not the source of plasticity. The deeper source comes from high-order nonlinear data forcing stacked piecewise manifolds to form ring / torus-like stationary structures.
These coupled stationary structures create elastic directions where weak perturbations can move the solution without destroying previously learned geometry.
In this sense, layer-wise isometry is an important preservation mechanism, while Deep Manifold places plasticity inside a broader geometric picture: node-cover reorientation, accumulated curvature, interconnected toroidal geometry, and eventual manifold rigidity.
#DeepManifoldInterpretation
📄 Paper 6/6
Preserving Plasticity in Continual Learning via Dynamical Isometry
Rosseau A, Müller R, Nowe A.
Accepted at ICML 2026 Main Track
ICML: icml.cc/virtual/2026/poster/…
@deepqlearning
1
4
43
3,360
📄 Paper 6/6
Preserving Plasticity in Continual Learning via Dynamical Isometry
Rosseau A, Müller R, Nowe A.
Accepted at ICML 2026 Main Track
ICML: icml.cc/virtual/2026/poster/…
@deepqlearning
1
5
3,548
📄 Paper 5/6
Reinforcement Learning for Tool-Calling Agents in Fast Healthcare Interoperability Resources (FHIR)
Knorr M*, Müller R*, Bremer JP, Schweingruber N.
Accepted at ICML 2026 Main Track
arXiv: arxiv.org/pdf/2605.14126
ICML: icml.cc/virtual/2026/poster/…
@deepqlearning
3
85
📄 Paper 3/6
Multi-Agent Reinforcement Learning of Karma Bidding Strategies
Riehl K, Psarou A, Müller R, Wu F, Langer P, Jakob R, Hollbeck G, Kouvelas A, Kucharski R, Makridis MA.
Accepted at ICML 2026 NExT-Game Workshop
OpenReview: openreview.net/forum?id=HaUU…
@DerRiehl @deepqlearning @PatrickLanger20 @robertjakob @gaborhollbeck
3
105
Excited to share that members of our lab co-authored 6 papers accepted at #ICML2026, including three Main Track and three Workshop papers 🔥🚀
📄 Accepted papers:
▪️ OpenTSLM: Time-Series Language Models for Reasoning over Multivariate Medical Text- and Time-Series Data [Main Track]
▪️ Auditing Emotion-Vector-Steered Political Bias in Open-Weight LLMs [AI4GOOD Workshop]
▪️ Reinforcement Learning of Karma Bidding Strategies [NExT-Game Workshop]
▪️ Cinematic Source Separation with Dialogue-Driven Sidechain Ducking [Workshop on Machine Learning for Audio]
▪️ Reinforcement Learning for Tool-Calling Agents in Fast Healthcare Interoperability Resources (FHIR) [Main Track]
▪️ Preserving Plasticity in Continual Learning via Dynamical Isometry [Main Track]
@robertjakob @PatrickLanger20 @gaborhollbeck @f14wn @DerRiehl @atoof_sh @deepqlearning @kev_osull @cs06thegreat @nzuma0 @maxrosenblattl
Huge congratulations to everyone involved. We are looking forward to presenting these works, reconnecting with colleagues, and meeting new friends in Seoul 🇰🇷
🔗 Full links in the comments.
5
7
19
1,779
Mar 25
【 Ren'Py 深度學習 (Deep Q-Learning)】
空師傅又來啦！
當 Galgame 引擎遇上神經網路，在 Ren'Py 做深度學習！
這AI走位是真的燒，
難道人工AI要被輾壓了嗎？(っ °Д °;)っ
youtu.be/A9WWtILHTp4
#遊戲開發 #renpy #遊戲引擎 #格鬥遊戲 #人工智慧 #強化學習 #深度學習 #deepqlearning
2
187
💥Highly recommended publication: "On Automated Object Grasping for Intelligent Prosthetic Hands Using Machine Learning"
🔗shorturl.at/WZY01
📌#DeepQLearning #AIinRobotics #ProstheticHands
1
5
59
24 Jul 2025
🔥 Read our Highly Cited Paper
📚 #DeepQLearning-Based Smart Scheduling of EVs for #DemandResponse in #SmartGrids
🔗 mdpi.com/2076-3417/14/4/1421
👨🔬 Viorica Rozina Chifu et al.
🏫 @utcluj
#EVscheduling #reinforcementlearning
3
49
18 Mar 2025
#DeepQlearning for continuous actions is key for controlling #robots, but it has been tricky to train #largeModels to get those performance gains. In these lectures I cover the fundamentals and explain how new research is bending the rules of #thedeadlytriad to advance #scaling.
1
22
106
5,491
30 Oct 2023
Deep Q Learning with PyTorch.
#PyTorch #development #python #deepqlearning
@ThePracticalDev @akshayballal95
dev.to/akshayballal/deep-q-l…
1
3
105
8 Apr 2023
Although training a pendulum to swing up is a difficult task in itself, the real challenge lies in defining the reward system to guide the learning process.
#screenshotsaturday #madewithunity #MachineLearning #DeepQLearning #PendulumProblems
6
62
12 Mar 2023
#MPO #Robotics #AI #ReinforcementLearning #RL #MachineLearning #LibTorch #PyTorch
#DeepQLearning algorithm in C that sees the #actorcritic and their targets as distributions and updates with a #KLDivergence.
Networks can be adjusted to make it #MOMPO
github.com/MotorCityCobra/C_…
3
130
4 Jan 2023
RT Applied Reinforcement Learning III: Deep Q-Networks (DQN) dlvr.it/SgNpVj #machinelearning #deepqlearning #dqn #artificialintelligence
1
1
117
7 Oct 2021
#DeepQLearning to Solve (Banana Collector). #BigData #Analytics #DataScience #AI #MachineLearning #ReinforcementLearning #IoT #IIoT #Python #RStats #TensorFlow #JavaScript #ReactJS #Serverless #DataScientist #Linux #Programming #Coding #100DaysofCode
bit.ly/3iB8Tzm
14
6
15 Sep 2021
Die Uhr läuft, in 60 Sekunden erklären wir @TMDTWuppertal , @Uni_Wuppertal, wieder einen #Forschungsbegriff. Diesmal: Was ist Deep Q-Learning? Das erklärt Euch @ja_r_pe auf unserem YouTube-Kanal
➡youtu.be/gN02oEdOzpI
#DeepQLearning #ReinforcementLearning #60secondspitches
1
5
13 Sep 2021
I accidentally commented out this line of code and it caused issues with my #AI training
Took me months to realize and now I get PTSD whenever I see it...
---
#100DaysOfCode #MachineLearning #NeuralNetwork #QLearning #DeepQLearning #ArtificialIntelligence
4
1