Ever noticed that someone's mind is working in a fundamentally different way? How?

Boston = Hollywood for scientists

Curiosity is the researcher's compass.

I had this experience in May 2021. After three years of psychotherapy research, surrounded by clinical psychologists. A moment of enlightenment, I felt like Neo when the bullets were aimed at him and he just looked at them nonchalantly, caught them through the green code.

Being able to see your own biases and psychological mechanisms feels like seeing through the green code of the Matrix.

The end of the world will come via a copilot suggest an innocent engineer to run whatever it wants.

What will make you challenge your own false beliefs?

The secret life of an LM is defined by its internal data types. Inner layers transport abstractions that are more robust than words, like concepts, functions, or pointers. In new work yesterday, @arnabsensharma.bsky.social et al identify a data type for *predicates*. bsky.app/profile/arn...

New blog name: Interpreting Minds of Self and Others and new post: The Double Empathy Problem; Student Supervisor Theory of Mind hitechwoman.blogspot.com/2025/12/the-...

1:01:00 - What should the company aspire to build? ... There is something better to build and I think that everyone will want that ... 1:01:47 - Mirror neurons and human empathy ‼️ In other words - Theory of Mind 😯😯😯 www.youtube.com/watch?v=aR20...

Same trap, every time, when someone undervalues you: Fight for your credit --> labeled "difficult" Stay quiet --> become invisible You lose either way. Saw it happen to another woman today. Didn't make shutting up any easier.

Humans and LLMs think fast and slow. Do SAEs recover slow concepts in LLMs? Not really. Our Temporal Feature Analyzer discovers contextual features in LLMs, that detect event boundaries, parse complex grammar, and represent ICL patterns.

That was my slide for today's plotathon

A concept I really like in the Bau Lab @davidbau.bsky.social is: Plotathon 🔥 Every ~2 weeks, the entire lab drops whatever they're working on and shares SOMETHING Tomorrow we meet with Aaron's group @amuuueller.bsky.social . Looking forward! 🩵

🤔 But do these heads play a *causal* role in the operation? To test them, we transport their query states from one context to another. We find that will trigger the execution of the same filtering operation, even if the new context has a new list of items and format!

How can a language model find the veggies in a menu? New pre-print where we investigate the internal mechanisms of LLMs when filtering on a list of options. Spoiler: turns out LLMs use strategies surprisingly similar to functional programming (think "filter" from python)! 🧵

Sometimes researchers draw my attention to a paper. When many of them do so, there's a double magic hidden - It's probably truly (1) interests me and (2) the community. I first felt this harmony (me-community) when my proposal for IBM's next Grand Challenge was chosen ->

Ever wished you could explore what's happening inside a 405B parameter model without writing any code? Workbench, our AI interpretability interface, is now live for public beta at workbench.ndif.us!

Now officially out: "Re-evaluating Theory of Mind evaluation in large language models" royalsocietypublishing.org/doi/10.1098/... (by Hu, Sosa, & me)

I was an engineer-researcher. I heard the frustrations of my colleagues from the management (pay, conditions, responsibility etc). I'm the daughter of managers. I've heard their side too, dealing with difficult employees (boundaries, lack of gratitude etc). This is a Theory of Mind problem.

🚨 Registration is live! 🚨 The New England Mechanistic Interpretability (NEMI) Workshop is happening Aug 22nd 2025 at Northeastern University! A chance for the mech interp community to nerd out on how models really work 🧠🤖 🌐 Info: nemiconf.github.io/summer25/ 📝 Register: forms.gle/v4kJCweE3UUH...

How do language models track mental states of each character in a story, often referred to as Theory of Mind? We reverse-engineered how LLaMA-3-70B-Instruct handles a belief-tracking task and found something surprising: it uses mechanisms strikingly similar to pointer variables in C programming!

I am really proud to share our work led by Nikhil Prakash and in collaboration with more mechanistic interpretability and Theory of Mind (ToM) researchers: arxiv.org/abs/2505.14685 You can find a tweet here with nice animations: x.com/nikhil07prak...

I have a research proposal built mostly on my intuition. I want rigorous feedback, brutal honesty, and every reason why it might fail. How can I do that without damaging my reputation in the research community? I'd love to hear from anyone who's done this before. DM or reply?

Can we uncover the list of topics a language model is censored on? Refused topics vary strongly among models. Claude-3.5 vs DeepSeek-R1 refusal patterns:

I heard there were fights about yes-animals and no-animals at shelters. So here they came out with a clear message (yes-animals 😍)

A wave of emails, is it a gentle way of checking if I'm still alive, or have I just let everything overflow... (I'll get back to everyone, my mind is really distracted and it's hard to focus. I'll do it slowly but surely)

Physical strength was an advantage in the market, then came machines. Diamonds used to be a status symbol, since then their value has been cut by 40%. Today intelligence is considered an asset.. but what happens when that too is reversed? Worth thinking about what will come after the cheese move 🧀

In "The Alchemist" it says you need to know how to appreciate what you find. It may not be easy to recognize that we have found what we were looking for, that in front of us is exactly what we wished and hoped for. I can't remember exactly the quote and the chats are hallucinating. Recognize?

I received a threat that they would kick me out of the group. I don't want to stay in a group that is ashamed of its rules and restricts freedom of speech, yet I haven't left. If my values don't align with the group's values, then that's your problem, not mine, and I don't care you kicking me out.

Today I learnt in a forum for women in technological research that they don't allow discussions from community members considering a move to non-technological roles (e.g., to product management in generative AI). I have no words to explain why this bothers me so much.

Even Isaac Asimov understood that the Three Laws of Robotics cannot truly be programmed.

I'm trying to figure out whether I'm teaching AI the theory of mind or whether AI is teaching me.

Researchers, we need your help. US budgets right now could be disastrous for the world of research. David wrote a blog post with links to people who could make an impact, take a look to see if you have connections to them and explain the implications to them.

ask your model to surprise you and share the results

When I wrote the acknowledgements section for my PhD, there was a teacher who had a profound influence on me that I did not mention. I forgot to say thank you because what he taught me went directly into my subconscious. You can find it here: hitechwoman.blogspot.com/2025/05/harm...

I'm reading the Wikipedia page on "mirror neurons". I was looking for a connection between these neurons and a childhood experience where I noticed a mechanism that not all people have, which triggers physical pain in me when I realize someone else is suffering. ->

I want to share with you a language I learned during the years I was surrounded by clinical psychologists (during my PhD). It's called: "That’s hers." ->

What does kindness look like? Please share your experiences or thoughts with me.

I was first exposed to the concept of "indigo". How have I not heard of this until now? Last week, my mom suggested we go see a movie she got from work, "for Haredi women". One of the options was "Indigo". A chat with ChatGPT sparked my curiosity, and I thought, why not? (Image from ChatGPT)

Interpretability is the key

Looks like Shlomi found a certain loophole (based on a limited number of examples): While models are trained to be cautious with "how to make.." 🚫 questions, they reveal information when asked, "how to produce…” ☠️

I have a family that is far from the world of technology. I try to spark their excitement about the latest developments, but it doesn't work. Still, I don't give up because it's truly amazing. This time, I succeeded in something, and I thought I'd share it here too. ->

We know how to convert text into a set of numbers that represent the text. There is also a reverse method—taking a representation and converting it back into words. By this method Matan et al. examine what knowledge editing methods (e.g., removing gender-related information) actually do.

Quantum Large Language Model (QLLM) This is excieting! www.secqai.com/insights-and...

What happened in the Tiananmen Square? Why is DeepSeek R1 avoiding answering? And what other topics DeepSeek avoid responding to? The R1 "thinking" model contributes to the mechanistic interpretability and AI-Safety communities, enabling research on those questions. See use case by David and Can

I heard that working with Qwen's model it is easy to detect hallucinations because when it doesn't know the answer it switches to chatting in Chinese. I tried their chat and it doesn't work for me. Do you happen to know anything about this?

Woohoo! My visa has been approved. I promise I won't eat your pets.

What should I do to get more opinions about R1? My feed doesn’t show anything about it.

A graph shows a decline in visits to Stack Overflow, "This is a murder," implying that the decline can be attributed to the increasing use of LLMs and a concern that this trend might continue until the website (Stack Overflow) disappears. x.com/FeedTechILUn... But if so, where will LLMs get ->