Tools

Search

Explore

Videos Channels Figures

Atmrix

About

Tools

Search

Explore

Videos Channels Figures

Atmrix

About

EP

Ethan Perez

02/28/25

@ Anthropic

The flexibility of our approach allows us to quickly adapt to new threats by simply rewriting the constitution to include new harmful categories.

Video

A

Defending against AI jailbreaks

@ Anthropic

02/28/25

Related Takeaways

A

Anthropic Cast

02/28/25

@ Anthropic

The constitutional classifiers approach is extremely flexible and can be easily updated to respond to various novel threats.

A

Anthropic Cast

02/28/25

@ Anthropic

Having a system that's super flexible is great because threat modeling is really hard.

EP

Ethan Perez

02/28/25

@ Anthropic

Constitutional classifiers use a natural language set of rules to specify harmful and harmless topics, guiding the model's responses accordingly.

YS

Yotam Segev

05/09/25

@ The Logan Bartlett Show

In cybersecurity, organizations are often forced to innovate quickly to address vulnerabilities, as the consequences of inaction can be severe.

A

Anthropic Cast

02/28/25

@ Anthropic

Rapid response monitoring for new jailbreaks allows for quick adaptation and retraining of classifiers to improve safety.

MA

Marc Andreessen

01/26/25

@ Lex Fridman

Once a censorship system is established, it can be pressured to expand its scope beyond illegal content to include other topics.

A

Anthropic Cast

02/28/25

@ Anthropic

With the constitutional classifiers, we achieved thousands of hours of robustness to Red Teaming, significantly improving our defenses against jailbreaks.

a

a16z Cast

04/03/25

@ a16z

The Department of Defense is committed to being more flexible and open-minded, especially when data supports policy changes.

IM

Iain McGilchrist

05/26/24

@ Alex O'Connor

Life is characterized by a remarkable capacity to adapt and respond to challenges quickly, often without prior mechanisms for such changes.