Benjamin Fuhrer

[Benjamin Fuhrer @BenjaminFuhrer]

I’ll be presenting GBRL next Wednesday at ICML this year.
Come chat!
#icml2025

[AK @_akhaliq]

Gradient Boosting Reinforcement Learning

Neural networks (NN) achieve remarkable results in various tasks, but lack key characteristics: interpretability, support for categorical features, and lightweight implementations suitable for edge devices. While ongoing efforts aim to address these challenges, Gradient Boosting Trees (GBT) inherently meet these requirements. As a result, GBTs have become the go-to method for supervised learning tasks in many real-world applications and competitions. However, their application in online learning scenarios, notably in reinforcement learning (RL), has been limited. In this work, we bridge this gap by introducing Gradient-Boosting RL (GBRL), a framework that extends the advantages of GBT to the RL domain. Using the GBRL framework, we implement various actor-critic algorithms and compare their performance with their NN counterparts. Inspired by shared backbones in NN we introduce a tree-sharing approach for policy and value functions with distinct learning rates, enhancing learning efficiency over millions of interactions. GBRL achieves competitive performance across a diverse array of tasks, excelling in domains with structured or categorical features.

[Benjamin Fuhrer @BenjaminFuhrer]

Heading to #ICML2025 to share our GBRL poster!
• Gradient-boosted trees for RL
• Strong performance & OOD robustness
• CUDA-fast, SB3-ready, lightweight deployment
Let’s talk in Vancouver!
📄 https://t.co/vh3fpKO4JT
💻 https://t.co/fL50iClyen
@DalalGal @ChenTessler #RL #AI

[Benjamin Fuhrer @BenjaminFuhrer]

@liimeleemon Congrats! Great work!

[Benjamin Fuhrer @BenjaminFuhrer]

@Yampeleg בוצע.

[Gal Dalal @DalalGal]

We released a multi-agent RL framework for network congestion control with the first public realistic network simulator! https://t.co/nzrj0gngNO. Based on the amazing work of @BenjaminFuhrer and @ChenTessler

[Gal Dalal @DalalGal]

Very excited to present the first #AI-based datacenter network congestion control solution that finally runs on live NICs and beats all other SOTA competitors! @BenjaminFuhrer @tesslerc @YShpigelman https://t.co/yf5xVAcyVf
#NVIDIA #reinforcementlearning #cloudnetworking

[Benjamin Fuhrer @BenjaminFuhrer]

I just published in @TDataScience Integer-Only Inference for Deep Learning in Native C https://t.co/YHoCLFvsWj