Identifiability is proven for recurrent nonlinear switching dynamical systems under flexible assumptions, and ΩSDS is introduced as a flow-based estimator that improves disentanglement and forecasting over VAE-based methods.
Behavioral and brain sciences , volume=
5 Pith papers cite this work. Polarity classification is still indexing.
5
Pith papers citing it
representative citing papers
LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
NEO induces compositional latent programs as world theories from observations and executes them to enable explanation-driven generalization.
V-JEPA models trained only on feature prediction from 2 million public videos achieve 81.9% on Kinetics-400, 72.2% on Something-Something-v2, and 77.9% on ImageNet-1K using frozen ViT-H/16 backbones.
Outcome-level RL with binary or composite rewards improves compositional generalization over supervised fine-tuning by avoiding overfitting to frequent training patterns.
citing papers explorer
-
End-to-End Identifiable and Consistent Recurrent Switching Dynamical Systems
Identifiability is proven for recurrent nonlinear switching dynamical systems under flexible assumptions, and ΩSDS is introduced as a flow-based estimator that improves disentanglement and forecasting over VAE-based methods.
-
Stories in Space: In-Context Learning Trajectories in Conceptual Belief Space
LLMs perform in-context learning as trajectories through a structured low-dimensional conceptual belief space, with the structure visible in both behavior and internal representations and causally manipulable via interventions.
-
Learning to Theorize the World from Observation
NEO induces compositional latent programs as world theories from observations and executes them to enable explanation-driven generalization.
-
Revisiting Feature Prediction for Learning Visual Representations from Video
V-JEPA models trained only on feature prediction from 2 million public videos achieve 81.9% on Kinetics-400, 72.2% on Something-Something-v2, and 77.9% on ImageNet-1K using frozen ViT-H/16 backbones.
-
Reinforcement Learning for Compositional Generalization with Outcome-Level Optimization
Outcome-level RL with binary or composite rewards improves compositional generalization over supervised fine-tuning by avoiding overfitting to frequent training patterns.