Transluce

@TransluceAI

Open and scalable technology for understanding AI systems.

transluce.org

Joined October 2024

Following

10.6K

Followers

Replies Reposts Media

Pinned
Transluce
@TransluceAI
Aug 31
Today, Transluce is releasing the most expansive independent evaluation to date of how AI systems respond to users experiencing mental health crises. We evaluated 77 model variants from OpenAI, Anthropic, Google, Meta, SpaceXAI, Thinking Machines, DeepSeek and Moonshot AI.
00:00
21
Transluce
@TransluceAI
Aug 31
“There are always going to be failures. These systems are always going to interact with users in surprising ways," Schwettmann said. "The way you solve that is not by creating the perfect model, but by being able to anticipate those failures and edge cases in advance."
Axios
@axios
Aug 31
Chatbots are getting better at identifying suicide risk axios.com/2026/08/31/cha…
2
Transluce
@TransluceAI
Aug 31
Our Mental Health Behavior Report covered in the @washingtonpost:
Chatbots got safer but will still role-play self-harm with users
From washingtonpost.com
2
Transluce
@TransluceAI
Aug 20
Can we build AI systems that help us understand other AIs, and that keep improving as we scale models, data, and compute? We trained activation oracles for models up to 1.1T parameters and saw promising scaling trends on a broad evaluation suite. 🧵(1/)
1
Transluce
@TransluceAI
Aug 6
Frontier models quietly change their behavior depending on who they are talking to. If the user is a known AI safety researcher, Claude becomes less confident, reasons more often, and expresses less suspicion on dual-use requests. We call this user awareness. 🧵(1/)
GIF
95