Kaiwen Sheng

The paper identifies directional elements in brain dynamics that encode causal control hierarchies and show robustness to neural state manipulations.

This paper introduces CANDY, a framework that aligns neural and behavioral data across sessions and subjects, enabling cross-session behavior decoding and identifying shared neural dynamics.

This paper introduces exponentiated gradient (EG) as a learning algorithm for artificial neural networks that better aligns with biological principles than gradient descent.

This paper introduces DeepDAM, a framework that improves neural connectivity inference by aligning synthetic and in vivo data domains, achieving high accuracy in rodent and other systems.

The paper introduces U-RISC, a high-resolution EM dataset for cell membrane segmentation, and evaluates deep learning methods, finding a gap between current performance and human-level accuracy.

This paper introduces a general LSTM-based deep learning method, GNE, for automated parameter tuning of neuronal models and neural circuits, showing strong performance on both synthetic and experimental data.