# Auditing LLMs and Twitter

Page: https://stenobird.com/podcast/data-skeptic/auditing-llms-and-twitter
Text version: https://stenobird.com/podcast/data-skeptic/auditing-llms-and-twitter.md
Podcast: [Data Skeptic](https://stenobird.com/podcast/data-skeptic)
Published: 2025-01-29T18:53:00+00:00
Episode link: https://dataskeptic.com/blog/episodes/2025/auditing-llms-and-twitter
Audio file: https://pscrb.fm/rss/p/mgln.ai/e/35/traffic.libsyn.com/secure/dataskeptic/auditing-llms-and-twitter.mp3?dest-id=201630
Processing state: failed
JSON: https://stenobird.com/v1/public/podcasts/data-skeptic/episodes/auditing-llms-and-twitter
Duration seconds: 2426

## Resource

Our guests, Erwan Le Merrer and Gilles Tredan, are long-time collaborators in graph theory and distributed systems. They share their expertise on applying graph-based approaches to understanding both large language model (LLM) hallucinations and shadow banning on social media platforms. In this episode, listeners will learn how graph structures and metrics can reveal patterns in algorithmic behavior and platform moderation practices. Key insights include the use of graph theory to evaluate LLM outputs, uncovering patterns in hallucinated graphs that might hint at the underlying structure and training data of the models, and applying epidemic models to analyze the uneven spread of shadow banning on Twitter. ------------------------------- Want to listen ad-free? Try our Graphs Course? Join Data Skeptic+ for $5 / month of $50 / year https://plus.dataskeptic.com

## Actions

- request_transcript: `POST https://stenobird.com/v1/public/podcasts/data-skeptic/episodes/auditing-llms-and-twitter/transcription-requests` — Idempotently request low-priority transcript generation for this episode.
- read_markdown: `GET https://stenobird.com/podcast/data-skeptic/auditing-llms-and-twitter.md` — Read the agent-friendly Markdown representation of this episode resource.

A page view does not enqueue transcription. Agents should invoke `request_transcript` explicitly when they need this episode processed.

## Transcript

Full transcripts are not published on public pages unless there is a clear rights basis.