Machine learning 9
- On evals
- Understanding mixture-of-depths
- Inductive bias of neural networks on 1D regression: an empirical examination
- Non-factorised identifiable variational autoencoders for causal discovery and out-of-distribution generalisation
- Are graph neural networks (GNNs) fundamentally bottlenecked?
- Sharpness-aware minimisation and label noise
- Comparing GPT-3 and RNNs as probabilistic generative models
- Investigating short-term climate forecasts with surrogate modelling
- Intelligent game playing with AI