Summary from listing
Ericsson Research is seeking a motivated student to conduct research on Latent Neural-Operator Structural Causal Models for physical-layer systems. The project involves reviewing causal representation learning and identifiability literature, building a controlled physical-layer simulator, implementing the proposed model and baselines, evaluating latent-structure identifiability and compositional generalisation, and presenting findings in regular discussions and a final thesis. The role requires a largely completed master’s degree, strong academic performance, knowledge of machine learning, linear algebra, and probability, Python proficiency, and experience with PyTorch or JAX.
