The Information Bottleneck principle and neural networks
Publications
Early motion processing circuit uses gap junctions to achieve efficient stimuli encoding
Siwei Wang, Noga Zaslavsky, Alexander Borst, Naftali Tishby, Idan Segev
2017
COSYNEEfficient encoding of motion is mediated by gap junctions in the fly visual system
Siwei Wang, Alexander Borst, Noga Zaslavsky, Naftali Tishby, Idan Segev
2017
PLoS Comput BiolDeterministic annealing and the evolution of Information Bottleneck representations
Noga Zaslavsky, Naftali Tishby
2019
Technical reportInformation-Theoretic Principles in the Evolution of Semantic Systems
Noga Zaslavsky
2020
PhD Thesis, The Hebrew UniversityToward human-like object naming in artificial neural systems
Tiwalayo Eisape, Roger Levy, Joshua Tenenbaum, Noga Zaslavsky
2020
BAICS @ ICLRTowards Human-Agent Communication via the Information Bottleneck Principle
Mycal Tucker, Julie Shah, Roger Levy, Noga Zaslavsky
2022
RSS Workshop on Social Intelligence in Humans and RobotsTrading off Utility, Informativeness, and Complexity in Emergent Communication
Mycal Tucker, Julie Shah, Roger Levy, Noga Zaslavsky
2022
NeurIPS