Reinforcement Learning Conference 2026

From the 15th to the 18th of August, Logan Mondal Bhamidipaty (PhD Student) attended the third Reinforcement Learning Conference (RLC) that took place at the University of Montreal and presented his ongoing work in two workshops:

• RENEW: Towards Learning World Models and Repairing Model Exploitation from Preferences at Finding the Frame Workshop
• Imperfect World Models are Exploitable at RL in Big Worlds Workshop

At the main conference, Assistax: A Multi-Agent Hardware-Accelerated Reinforcement Learning Benchmark for Assistive Robotics, was published, with the work led by Leonard Hinckeldey (PhD student) with collaborators from Honda Research Institute EU, developing a Python library that provides multi-agent hardware-accelerated reinforcement learning environments in the domain of assistive robotics, together with accompanying baseline algorithm implementations, utilising JAX and Brax for quick training pipelines. Asistax received an outstanding paper award for tooling, environments, and evaluation for reinforcement learning.

More about RLC