Adnan KhanAdnanProduct & AI
San Francisco Bay Area
Get in touch

Source note2 min read

Themis: An explainable AI-enabled framework for Reinforcement Learning with Human Feedback

Source

arXiv

arxiv.org/abs/2606.24622v1

Read the source

Source

Plain-Language Brief

Researchers present Themis, an AI framework that combines explainability and human feedback to train safe Reinforcement Learning (RL) systems. The framework is designed for over 200 environments, easy configuration for RL, transparency, and alignment experiments. It outperforms the environment's true reward signal using human preferences and supports large participant groups in cloud-based experiments.

What the Document Says

  • Themis is an AI framework for Reinforcement Learning (RL) that focuses on transparency and alignment via human feedback.
  • It works with over 200 environments and is configurable for various RL, transparency, and alignment experiments.
  • Themis can train reward models that match or outperform the environment's true reward signal using human preferences.
  • There is a cloud-based platform for collecting human feedback and managing experiments which is user-friendly, auto-scalable, and supports large participant groups without extra development overhead.
  • Tests show Themis can support one thousand users in back-to-back experiments on a modest commercial machine.

Key Concepts Mentioned

  • Reinforcement Learning
  • Themis Framework
  • Transparency
  • Alignment
  • Human Feedback
  • Reward Models

Claims Made by the Document

  • Themis can train reward models that match or outperform the environment's true reward signal using human preferences.

Evidence Used

  • Results from experiments with Themis are presented in the text.

Methods / Approach

UNKNOWN

Limitations Stated by the Authors

UNKNOWN

Limitations Noted by the Agent

Agent interpretation — not from the source.

UNKNOWN

Open Questions

UNKNOWN

Notes for Future Q&A

May include interpretation. Labeled where it does.

UNKNOWN

Raw Abstract / Excerpt

Training safe Reinforcement Learning (RL) systems is inherently challenging, with no guarantee of avoiding unwanted behaviors. The most effective defenses against this are (i) transparency through explainability and (ii) alignment via human feedback. While both show promising results, no publicly available framework currently combines them. To address this, we introduce Themis, an XAI-enabled testing and evaluation framework for Reinforcement Learning from Human Feedback. Themis supports over 200 widely used environments and is easily configurable for experiments in RL, transparency, and alignment. Our results show that Themis can train reward models that match or outperform the environment's true reward signal using human preferences. We also provide a cloud-based platform for collecting human feedback and managing experiments. It is user-friendly, auto-scalable, and supports large participant groups across multiple experiments without extra development overhead. Tests show Themis can support one thousand users in back-to-back experiments on a modest commercial machine.