Workshop on Actionable Interpretability for Language Models and Machine Translation Systems

As large language models (LLMs) increasingly power high-stakes language technologies, understanding how they make decisions has never been more important. This workshop brings together researchers focused on interpretability and bias in AI systems, especially within machine translation and natural language processing. We will explore how actionable insights from interpretability research can drive practical improvements in reliability, reduce biases, mitigate hallucinations, and elevate overall output quality.

By fostering collaboration between interpretability researchers and AI practitioners, the workshop aims to accelerate the adoption of methods that make modern language systems more transparent, trustworthy, and controllable.

Who should attend?
Researchers and practitioners in AI, computational linguistics, and related fields, as well as domain experts, particularly translation professionals, who work with language models in their workflows.