We've opened a new Postdoctoral Fellowship on *LLM-Based Reasoning and Planning* jointly with Nada Amin’s group Harvard SEAS Interested in developing fundamental methods in AI to help doctors and civic policymakers make more informed decisions? Learn more: bit.ly/3wAyvWQ

We're hiring for a new role: Research Operations Assistant. Work with our research teams to grow our operations and increase our impact! Details here: basis.ai/roles/research… RTs much appreciated!

New ARC-AGI paper ARC Prize w/ fantastic collaborators Wen-Ding Li @ ICLR'25 Keya Hu Zenna Tavares evanthebouncy Basis For few-shot learning: better to construct a symbolic hypothesis/program, or have a neural net do it all, ala in-context learning? cs.cornell.edu/~ellisk/docume…

Looking forward to verifying this one. Kevin Ellis and team are some of the best folks today working on program synthesis.

Basis has started a joint project called MARA w/ Kevin Ellis 's lab +others. This is our first output. More details soon. We're sponsoring "Systems 2 Reasoning at Scale" workshop at Neurips & will present MARA there. We're hiring for it now! basis.ai/join-us/#caree…

New work with Kevin Ellis and collabs on ARC Prize ARC-AGI, including new SOTA results.

🎡🦆New Basis paper by Rafal Urbaniak, Marjorie Xie and me! Linking cognitive strategy, neural mechanism, and movement statistics in group foraging behaviors nature.com/articles/s4159… "Behind the paper" blog here: go.nature.com/3YBaVDK Code: github.com/BasisResearch/… 🦆🎡

📢🎡🐦‍⬛I'm looking to hire postdocs to join me on the Collaborative Intelligent Systems project at Basis, more info here: basis.ai/roles/postdoc-…, please apply 🐦‍⬛🎡📢

🚀We’re hiring! Forest Neurotech is looking for a Software Engineering Lead to build the core systems powering our ultrasound neurotech platform. As a nonprofit FRO, we're advancing science for public good. If you’re excited about neurotech & impact let’s talk. 🌍🧠 Link below

Proud to share that our work with Kevin Ellis and collabs won the 1st prize ARC Paper Award! This is the first work to come out of the MARA project. Much more to come.

My big list of ARC Prize 2024 surprises: 1. TTT works really well to solve "novel" problems. Assuming you have a way to do data augmentation on the fly. 2. Brute force program synthesis is competitive with frontier LLM/AI approaches. We'll fix this in v2. 3. The private and

Thrilled that joint work by Kevin Ellis's lab and Basis won 1st prize in ARC Prize Paper Awards and 2nd prize in ARC-AGI-PUB (w/ MIT) This is our first result from Project MARA: an effort to build Modeling, Abstraction, and Reasoning Agents capable of "everyday science"

Thank you, François, Mike, & team, for the ARC challenge. It has been a durable source of inspiration, and brings fresh ideas to AI. The paper award first authors are Keya Hu (applying to PhDs Keya Hu) and Wen-Ding Li (at NeurIPS hunting for industry gigs Wen-Ding Li @ ICLR'25).

Chollet François Chollet : [paraphrase] "using NNs for discrete problems i.e. finding new prime numbers - is a terrible idea" with Kevin Ellis Zenna Tavares

Clip of François Chollet reflecting on the ARC Prize challenge with Zenna Tavares and Kevin Ellis at #neurips2024

This was a lot of fun.

We’re making the Basis organisation document public today. It’s less a charter and more a design doc—our spec for why a new kind of technology and research organisation is needed, and how to build it. (link below)

Thrilled to see our TinyRNN paper in @nature! We show how tiny RNNs predict choices of individual subjects accurately while staying fully interpretable. This approach can transform how we model cognitive processes in both healthy and disordered decisions. doi.org/10.1038/s41586…

New paper from Basis' Project MARA team and collabs. The ability to learn and use world models is a key aspect of human intelligence, but evaluating this ability remains elusive. In this work we propose WorldTest, a representation-agnostic, behavior-based agent eval framework.