![[Paper] Difficulties with Evaluating a Deception Detector for AIs](https://unco.digital/wp-content/uploads/2025/10/139787-1.jpg)
[Paper] Difficulties with Evaluating a Deception Detector for AIs
Published on December 3, 2025 8:07 PM GMTNew research from the GDM mechanistic interpretability team.…
Recent Articles
View All
MIT engineers design an aerial microrobot that can fly as fast as a bumblebee
With insect-like speed and agility, the tiny robot could someday aid in search-and-rescue missions.
6 reasons why “alignment-is-hard” discourse seems alien to human intuitions, and vice-versa
Published on December 3, 2025 6:37 PM GMTTl;drAI alignment has a culture clash. On one…
OpenAI has trained its LLM to confess to bad behavior
OpenAI is testing another new way to expose the complicated processes at work inside large…
Anthropic hires lawyers as it preps for IPO
Anthropic is reportedly prepping for an IPO that could come as early as 2026.
The Complete Guide to Using Pydantic for Validating LLM Outputs
Large language models generate text, not structured data.
AWS doubles down on custom LLMs with features meant to simplify model creation
AWS is launching more capabilities in both Amazon Bedrock and Amazon SageMaker AI to make…
Another bid to block state AI regulation has failed…for now
Republicans’ attempt to ban state AI regulations was removed from the defense bill after bipartisan…
Look back on your 2025 with Google Photos Recap
Learn more about Google Photos Recap — now available for 2025 — and how you…
Helping power-system planners prepare for an unknown future
Macro, a modeling tool developed by the MIT Energy Initiative, enables energy-system planners to explore…