ML Safety Newsletter

ML Safety Newsletter #14

Resisting Prompt Injection, Evaluating Cyberattack Capabilities, and SafeBench Winners

May 7 • 

Alice Blair

 and 

Dan Hendrycks

April 2025

ML Safety Newsletter #13

Chain-of-Thought Monitoring, Distinguishing Honesty from Accuracy, and Emergent Misalignment

Apr 2 • 

Julius Simonelli

 and 

Dan Hendrycks

February 2025

ML Safety Newsletter #12: February 2025

ML Safety Newsletter Relaunch

Feb 27 • 

Julius Simonelli

 and 

Dan Hendrycks

December 2023

ML Safety Newsletter #11

Top Safety Papers of 2023

Dec 14, 2023 • 

Dan Hendrycks

 and 

Aidan O'Gara

September 2023

ML Safety Newsletter #10

Adversarial attacks against language and vision models, improving LLM honesty, and tracing the influence of LLM training data

Sep 13, 2023 • 

Dan Hendrycks

 and 

Aidan O'Gara

April 2023

ML Safety Newsletter #9

Verifying large training runs, security risks from LLM access to APIs, why natural selection may favor AIs over humans

Apr 11, 2023 • 

Dan Hendrycks

 and 

Thomas Woodside

February 2023

ML Safety Newsletter #8

Interpretability, using law to inform AI alignment, scaling laws for proxy gaming

Feb 20, 2023 • 

Dan Hendrycks

 and 

Thomas Woodside

January 2023

ML Safety Newsletter #7

Making model dishonesty harder, making grokking more interpretable, an example of an emergent internal optimizer

Jan 9, 2023 • 

Dan Hendrycks

October 2022

ML Safety Newsletter #6

Transparency survey, provable robustness, models that predict the future

Oct 13, 2022 • 

Dan Hendrycks

September 2022

ML Safety Newsletter #5

Safety competitions with more than $1 million in prizes

Sep 26, 2022 • 

Dan Hendrycks

June 2022

ML Safety Newsletter #4

Many New Interpretability Papers, Virtual Logit Matching, Rationalization Helps Robustness

Jun 3, 2022 • 

Dan Hendrycks

March 2022

ML Safety Newsletter #3

Transformer adversarial robustness, fractals, preference learning

Mar 8, 2022 • 

Dan Hendrycks