← All episodes
EPISODE 477

Why AI Decision Explanations Are Broken

Learning GenAI via SOTA Papers · October 8, 2026 · 25 min

This paper establishes a formal mathematical and causal framework for tracing, estimating, and attributing decisions within dynamic agentic systems. It provides essential foundational theoretical primitives required to formally audit, verify, and explain complex autonomous agent reasoning loops.

Short on time? Watch the video overview of this paper.

Listen on: Spotify · Apple Podcasts · Amazon Music · YouTube

Follow the show: YouTube · Medium · X · LinkedIn · Bluesky · Threads · Instagram · TikTok · Facebook · Telegram