tweetindex
IT

Transluce ✓

@TransluceAI · joined 12 Oct 2024

Open and scalable technology for understanding AI systems.

10 660Followers
21Following
345Posts total
1.3MViews on collected posts

Ultimi post

“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." 1.7K views · 16 likes · 0 reposts · 2 replies Open on X →
Our Mental Health Behavior Report covered in the @washingtonpost: https://t.co/kTedG1MFev 1.3K views · 16 likes · 1 reposts · 2 replies Open on X →
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. https 67K views · 425 likes · 82 reposts · 21 replies Open on X →
Chatbots are getting better at identifying suicide risk https://t.co/sA6TFfwYbX 24.8K views · 10 likes · 1 reposts · 4 replies Open on X →
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/) https://t.co/roqovGQf4i 13.7K views · 212 likes · 37 reposts · 1 replies Open on X →
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/) https://t.co 779.5K views · 2.8K likes · 298 reposts · 95 replies Open on X →
@TransluceAI Excited to look into this. I’ve taken a lot of inspiration from Sarah’s work on MAIA and FIND! Happy to see this nonprofit come about. Cc @BogdanIonutCir2 1.3K views · 6 likes · 0 reposts · 0 replies Open on X →
To learn more and get involved, check out our Company page for how to contribute to our open-source tools: https://t.co/VAobVegMkK. 2.6K views · 20 likes · 0 reposts · 0 replies Open on X →
The team of AI explorers can exploit the vastness of the cavern as a source of information for training and improvement, putting us on the right side of the bitter lesson. 2.7K views · 21 likes · 0 reposts · 1 replies Open on X →
As a non-profit, we can be a convening point for discussing best practices for understanding and verifying AI. We will present ideas, get feedback from the community, and iterate. We will build positive-sum partnerships with companies and governments, and help ensure that 3K views · 20 likes · 0 reposts · 1 replies Open on X →
In the long run, our vision is to create AI-driven tools that direct massive computational power toward explaining these complex systems. We imagine a human trying to understand an AI system as an explorer situated in a vast cavern. We want teams of agents to map all the 3.9K views · 25 likes · 1 reposts · 1 replies Open on X →
We’ve only been around a short time, but have already built several pieces towards our vision of scalable, AI-backed tools for understanding AI systems. You can read more about them in our release announcement: https://t.co/Iq5rwughV6. We’ll also share more throughout the day. 3.5K views · 26 likes · 0 reposts · 1 replies Open on X →
Transluce exists to fill this need. We are a non-profit, but operate at start-up speeds, building scalable, usable tech backed by AI. 3.7K views · 24 likes · 0 reposts · 1 replies Open on X →
AI will have extraordinary consequences for society. Given its public impact, the tools for understanding and verifying its safety must be open and publicly vetted. 4.5K views · 28 likes · 0 reposts · 1 replies Open on X →
Announcing Transluce, a nonprofit research lab building open source, scalable technology for understanding AI systems and steering them in the public interest. Read a letter from the co-founders Jacob Steinhardt and Sarah Schwettmann: https://t.co/IUIhBjpYhS https://t.co/jl91xv 347K views · 685 likes · 149 reposts · 37 replies Open on X →

Rispetto ad account della stessa dimensione

5 post degli ultimi 90 giorni, accanto alla fascia di 10K–100K follower. arriva a molti, ma pochi di loro reagiscono.

Visualizzazioni mediane13 692questo account924mediana per 10K–100K
Copertura, %128.44%questo account3.62%mediana per 10K–100K
Interazione, %1.08%questo account1.52%mediana per 10K–100K
MetricaQuesto accountMediana per 10K–100KRapporto
Visualizzazioni mediane per post13 69292414.8×
Copertura (visualizzazioni ÷ follower)128.44%3.62%35.5×
Tasso di interazione1.08%1.52%0.71×

Altri account di questa fascia →   Confronta con un altro account →   Come sono costruiti questi parametri →

Growth & engagement

How the posts we collected actually performed: views and reaction rate post by post, what the audience did with them, and where the follower count goes.

Views per post

4.5K23 Oct
3.7K
3.5K
3.9K
3K
2.7K
2.6K
1.3K
779.5K6 Aug
13.7K20 Aug
24.8K31 Aug
67K
1.3K
1.7K

Last 14 collected posts, oldest on the left. The scale is logarithmic: one post can outrun the rest a hundred times over.

Engagement rate per post

0.65%23 Oct
0.68%
0.78%
0.77%
0.71%
0.82%
0.76%
0.46%
0.43%6 Aug
1.86%20 Aug
0.06%31 Aug
0.83%
1.42%
1.08%

Reactions — likes, reposts, replies and quotes — divided by views. Median for 10K–100K accounts is 1.52%.

What the audience does

Likes57.8%4 370 in total
Reposts7.5%569 in total
Replies2.2%168 in total
Quotes3.1%233 in total
Bookmarks29.4%2 224 in total

Share of every reaction we collected for this account. Replies mean argument, reposts mean endorsement, bookmarks mean the post was worth keeping.

The follower curve appears once this account has two daily snapshots — we take one a day, and this one is on its first.

Account simili