If you were to survey researchers, and ask them to name the 5 most broadly influential ideas in Machine Learning from the last 5 years, I’d bet good money that Batch Normalization would be somewhere on everyone’s lists. Before Batch Norm, training meaningfully deep neural networks was an unstable process, and one that often took a long time to converge to success. When we added Batch Norm to models, it allowed us to increase our learning rates substantially (leading to quicker training) without the risk of activations either collapsing or blowing up in values. It had this effect because it addressed one of the key difficulties of deep networks: internal covariate shift. 

To understand this, imagine the smaller problem, of a one-layer model that’s trying to classify based on a set of input features. Now, imagine that, over the course of training, the input distribution of features moved around, so that, perhaps, a value that was at the 70th percentile of the data distribution initially is now at the 30th. We have an obvious intuition that this would make the model quite hard to train, because it would learn some mapping between feature values and class at the beginning of training, but that would become invalid by the end. This is, fundamentally, the problem faced by higher layers of deep networks, since, if the distribution of activations in a lower layer changed even by a small amount, that can cause a “butterfly effect” style outcome, where the activation distributions of higher layers change more dramatically. Batch Normalization - which takes each feature “channel” a network learns, and normalizes [normalize = subtract mean, divide by variance] it by the mean and variance of that feature over spatial locations and over all the observations in a given batch - helps solve this problem because it ensures that, throughout the course of training, the distribution of inputs that a given layer sees stays roughly constant, no matter what the lower layers get up to. 

On the whole, Batch Norm has been wildly successful at stabilizing training, and is now canonized - along with the likes of ReLU and Dropout - as one of the default sensible training procedures for any given network. However, it does have its difficulties and downsides. One salient one of these comes about when you train using very small batch sizes - in the range of 2-16 examples per batch. Under these circumstance, the mean and variance calculated off of that batch are noisy and high variance (for the general reason that statistics calculated off of small sample sizes are noisy and high variance), which takes away from the stability that Batch Norm is trying to provide. 

One proposed alternative to Batch Norm, that didn’t run into this problem of small sample sizes, is Layer Normalization. This operates under the assumption that the activations of all feature “channels” within a given layer hopefully have roughly similar distributions, and, so, you an normalize all of them by taking the aggregate mean over all channels, *for a given observation*, and use that as the mean and variance you normalize by. Because there are typically many channels in a given layer, this means that you have many “samples” that go into the mean and variance. However, this assumption - that the distributions for each feature channel are roughly the same - can be an incorrect one. 

A useful model I have for thinking about the distinction between these two approaches is the idea that both are calculating approximations of an underlying abstract notion: the in-the-limit mean and variance of a single feature channel, at a given point in time. Batch Normalization is an approximation of that insofar as it only has a small sample of points to work with, and so its estimate will tend to be high variance. Layer Normalization is an approximation insofar as it makes the assumption that feature distributions are aligned across channels: if this turns out not to be the case, individual channels will have normalizations that are biased, due to being pulled towards the mean and variance calculated over an aggregate of channels that are different than them. 

Group Norm tries to find a balance point between these two approaches, one that uses multiple channels, and normalizes within a given instance (to avoid the problems of small batch size), but, instead of calculating the mean and variance over all channels, calculates them over a group of channels that represents a subset. The inspiration for this idea comes from the fact that, in old school computer vision, it was typical to have parts of your feature vector that - for example - represented a histogram of some value (say: localized contrast) over the image. Since these multiple values all corresponded to a larger shared “group” feature. If a group of features all represent a similar idea, then their distributions will be more likely to be aligned, and therefore you have less of the bias issue. 

One confusing element of this paper for me was that the motivation part of the paper strongly implied that the reason group norm is sensible is that you are able to combine statistically dependent channels into a group together. However, as far as I an tell, there’s no actually clustering or similarity analysis of channels that is done to place certain channels into certain groups; it’s just done so semi-randomly based on the index location within the feature channel vector. So, under this implementation, it seems like the benefits of group norm are less because of any explicit seeking out of dependant channels, and more that just having fewer channels in each group means that each individual channel makes up more of the weight in its group, which does something to reduce the bias effect anyway.

The upshot of the Group Norm paper, results-wise, is that Group Norm performs better than both Batch Norm and Layer Norm at very low batch sizes. This is useful if you’re training on very dense data (e.g. high res video), where it might be difficult to store more than a few observations in memory at a time. However, once you get to batch sizes of ~24, Batch Norm starts to do better, presumably since that’s a large enough sample size to reduce variance, and you get to the point where the variance of BN is preferable to the bias of GN.  

This paper explores the use of convolutional (PixelCNN) and recurrent units (PixelRNN) for modeling the distribution of images, in the framework of autoregression distribution estimation. In this framework, the input distribution $p(x)$ is factorized into a product of conditionals $\Pi p(x_i | x_i-1)$. Previous work has shown that very good models can be obtained by using a neural network parametrization of the conditionals (e.g. see our work on NADE \cite{journals/jmlr/LarochelleM11}). Moreover, unlike other approaches based on latent stochastic units that are directed or undirected, the autoregressive approach is able to compute log-probabilities tractably. 

So in this paper, by considering the specific case of x being an image, they exploit the topology of pixels and investigate appropriate architectures for this.

Among the paper's contributions are:

1. They propose Diagonal BiLSTM units for the PixelRNN, which are efficient (thanks to the use of convolutions) while making it possible to, in effect, condition a pixel's distribution on all the pixels above it (see Figure 2 for an illustration).

2. They demonstrate that the use of residual connections (a form of skip connections, from hidden layer i-1 to layer $i+1$) are very effective at learning very deep distribution estimators (they go as deep as 12 layers).

3. They show that it is possible to successfully model the distribution over the pixel intensities (effectively an integer between 0 and 255) using a softmax of 256 units.

4. They propose a multi-scale extension of their model, that they apply to larger 64x64 images.

The experiments show that the PixelRNN model based on Diagonal BiLSTM units achieves state-of-the-art performance on the binarized MNIST benchmark, in terms of log-likelihood. They also report excellent log-likelihood on the CIFAR-10 dataset, comparing to previous work based on real-valued density models. Finally, they show that their model is able to generate high quality image samples.

This paper deals with the question what / how exactly CNNs learn, considering the fact that they usually have more trainable parameters than data points on which they are trained.

When the authors write "deep neural networks", they are talking about Inception V3, AlexNet and MLPs.

## Key contributions

* Deep neural networks easily fit random labels (achieving a training error of 0 and a test error which is just randomly guessing labels as expected). $\Rightarrow$Those architectures can simply brute-force memorize the training data.
* Deep neural networks fit random images (e.g. Gaussian noise) with 0 training error. The authors conclude that VC-dimension / Rademacher complexity, and uniform stability are bad explanations for generalization capabilities of neural networks
* The authors give a construction for a 2-layer network with $p = 2n+d$ parameters - where $n$ is the number of samples and $d$ is the dimension of each sample - which can easily fit any labeling. (Finite sample expressivity). See section 4.


## What I learned

* Any measure $m$ of the generalization capability of classifiers $H$ should take the percentage of corrupted labels ($p_c \in [0, 1]$, where $p_c =0$ is a perfect labeling and $p_c=1$ is totally random) into account: If $p_c = 1$, then $m()$ should be 0, too, as it is impossible to learn something meaningful with totally random labels.
* We seem to have built models which work well on image data in general, but not "natural" / meaningful images as we thought.


## Funny

> deep neural nets remain mysterious for many reasons

> Note that this is not exactly simple as the kernel matrix requires 30GB to store in memory. Nonetheless, this system can be solved in under 3 minutes in on a commodity workstation with 24 cores and 256 GB of RAM with a conventional LAPACK call.


## See also

* [Deep Nets Don't Learn Via Memorization](https://openreview.net/pdf?id=rJv6ZgHYg)

This paper shows exciting results on using Model-based RL for Atari. 

Model-based RL has shown impressive improvements in sample efficiency on Mujoco tasks ([Chua et. al, 2018](https://arxiv.org/abs/1805.12114)), so its nice to see that the sample efficiency improvements carry over to Pixel-based envs like Atari too. 

Specifically, the authors show that their model-based method can do well on several Atari games after training on only 100K env steps (400K frames with FrameSkip 4) which roughly corresponds to 2 hours of game play. They compare to SOTA model-free variants (Rainbow, PPO) after similar number of frames and show that the model-based version achieves much better scores. 

The overall training procedure has a very Dyna like flavor. The algorithm, termed SimPLe follows an iterative scheme of:
* Collect experience from the real environment using a policy (initialized to random). 
* Use this experience to train the world model (a next-step frame prediction model, and a reward prediction model). This amounts to supervised learning on `{(s, a) -> s’}` and `{(s, a) -> r}` pairs. 
* Generate rollouts using the world model, and learn a policy with these rollouts using PPO.

https://i.imgur.com/SZLmdME.png

**Countering distributional shift:**

A key issue when training models is compounding errors when doing multi-step rollouts. This is similar to the problem of making predictions with RNNs trained via teacher-forcing, and hence it's natural to leverage existing techniques from that literature. 

This paper uses one such technique: scheduled sampling, that is during training randomly replace some frames of the input by the prediction from the previous step. This seems like a natural way to make the model robust to slight distributional changes. 


**Commentary / possible future work:** 

* The paper evaluated only on 26 out of 60 Atari games in ALE. I would have really liked if the authors showed performance numbers on all the games even if they weren’t good.  
* Related: I suspect the method would not work well when the initial diversity of frames given by the random policy is not sufficient (ex. Sparse reward games like Montezuma’s revenge/Pitfall). Using sample efficient exploration algorithms to augment model learning would be really interesting. 
* The trained world-model is able to rollout only for 50 time-steps (compounding errors don't allow for longer rollouts), it might be worthwhile to explore models that can do long-horizon predictions [(TD-VAE?)](https://openreview.net/forum?id=S1x4ghC9tQ). 
* Apart from sample-efficiency gains, one reason I am excited about models is their potential ability to generalize to different tasks in the same environment. Benchmarking their generalization capability should thus be an exciting next step. 

Finally, props to authors for open-sourcing the code: [tensor2tensor/tensor2tensor/rl at master · tensorflow/tensor2tensor · GitHub](https://github.com/tensorflow/tensor2tensor/tree/master/tensor2tensor/rl) and providing detailed instructions to run. 

This is a nice little empirical paper that does some investigation into which features get learned during the course of neural network training. To look at this, it uses a notion of "decodability", defined as the accuracy to which you can train a linear model to predict a given conceptual feature on top of the activations/learned features at a particular layer. This idea captures the amount of information about a conceptual feature that can be extracted from a given set of activations. 

They work with two synthetic datasets. 

1. Trifeature: Generated images with a color, shape, and texture, which can be engineered to be either entirely uncorrelated or correlated with each other to varying degrees. 
2. Navon: Generated images that are letters on the level of shape, and are also composed of letters on the level of texture 

The first thing the authors investigate is: to what extent are the different properties of these images decodable from their representations, and how does that change during training? In general, decodability is highest in lower layers, and lowest in higher layers, which makes sense from the perspective of the Information Processing Inequality, since all the information is present in the pixels, and can only be lost in the course of training, not gained. They find that decodability of color is high, even in the later layers untrained networks, and that the decodability of texture and shape, while much less high, is still above chance. When the network is trained to predict one of the three features attached to an image, you see the decodability of that feature go up (as expected), but you also see the decodability of the other features go down, suggesting that training doesn't just involve amplifying predictive features, but also suppressing unpredictive ones. This effect is strongest in the Trifeature case when training for shape or color; when training for texture, the dampening effect on color is strong, but on shape is less pronounced. 

https://i.imgur.com/o45KHOM.png

The authors also performed some experiments on cases where features are engineered to be correlated to various degrees, to see which of the predictive features the network will represent more strongly. In the case where two features are perfectly correlated (and thus both perfectly predict the label), the network will focus decoding power on whichever feature had highest decodability in the untrained network, and, interestingly, will reduce decodability of the other feature (not just have it be lower than the chosen feature, but decrease it in the course of training), even though it is equally as predictive. 
https://i.imgur.com/NFx0h8b.png

Similarly, the network will choose the "easy" feature (the one more easily decodable at the beginning of training) even if there's another feature that is slightly *more* predictive available. This seems quite consistent with the results of another recent paper, Shah et al, on the Pitfalls of Simplicity Bias in neural networks. The overall message of both of these experiments is that networks generally 'put all their eggs in one basket,' so to speak, rather than splitting representational power across multiple features. 

There were a few other experiments in the paper, and I'd recommend reading it in full - it's quite well written - but I think those convey most of the key insights for me.