with @robmok.bsky.social and Xiaoliang "Ken" Luo

Intuitive cell types don't necessarily play the ascribed functional role in the overall computation. This is not a message the field wants to hear as it suggests better baselines, controls, and some reflection. elifesciences.org/reviewed-pre... 2/2

Working with monkey data, we found neural representations stretched across brain regions to emphasize task relevant features on a trial-by-trial basis. Spike timing mattered over spike rate. Deep nets did the same. nature.com/articles/s41... 2/2

We developed a straightforward method of combining confidence-weighted judgments for any number of humans and AIs. w Felipe Yáñez, Omar Valerio Minero, @ken-lxl.bsky.social 2/2

[email protected], send to me, or send directly to the Met (London police) who are investigating www.met.police.uk. I could see this being super distressing for a vulnerable person, so hope this does not become more common. For me, it's been an exercise in rapidly learning to not care! 2/2

Bonus: I found it counterintuitive that (in theory) the learning problem is the same for any word ordering. Aligning proof and simulation was key. Now, new avenues open to address positional biases, better training and knowing when to trust LLMs. w @ken-lxl.bsky.social arxiv.org/abs/2505.08739

When LLMs diverge from one another because of word order (data factorization), it indicates their probability distributions are inconsistent, which is a red flag (not trustworthy). We trace deviations to self-attention positional and locality biases. 2/2 arxiv.org/abs/2505.08739

with @ken-lxl.bsky.social , @robmok.bsky.social , Brett Roads

A 7B is small enough to train efficiently on 4 A100s (thanks Microsoft) and at the time Mistral performed relatively well for its size.

Yes, the model weights and all materials are openly available. We really want to offer easy to use tools people can use through the web without hassle. To do that, we need to do more work (will be announcing an open source effort soon) and need some funding for hosting a model endpoint.

While BrainBench focused on neuroscience, our approach is science general, so others can adopt our template. Everything is open weight and open source. Thanks to the entire team and the expert participants. Sign up for news at braingpt.org 8/8

Finally, LLMs can be augmented with neuroscience knowledge for better performance. We tuned Mistral on 20 years of the neuroscience literature using LoRA. The tuned model, which we refer to as BrainGPT, performed better on BrainBench. 7/8

Indeed, follow-up work on teaming finds that joint LLM and human teams outperform either alone, because LLMs and humans make different types of errors. We offer a simple method to combine confidence-weighted judgements.
arxiv.org/abs/2408.08083 6/8