jacobfa/Attractor

Code for Bridging the Gap Between Latent and Explicit Reasoning with Looped Transformers (https://arxiv.org/abs/2606.31779)

Stable Looped Models and their Scaling Laws

The Truth Is In There: Improving Reasoning in Language Models with Layer-Selective Rank Reduction

[ICML 2026] Code for Equilibrium Reasoners: learning attractor dynamics for scalable reasoning

Code for COLM '26 paper 'Loop, Think, & Generalize: Implicit Reasoning in Recurrent-Depth Transformers'

Code for Quiet-STaR

Tiny Model, Big Logic: Diversity-Driven Optimization Elicits Large-Model Reasoning Ability in VibeThinker-1.5B