Recursive Sketches for Modular Deep Learning
This addresses interpretability and analysis challenges for researchers and practitioners working with modular deep learning systems, though it appears incremental as it builds on existing sketching techniques.
The paper tackles the problem of understanding complex modular deep networks by introducing recursive sketches that summarize input-output relationships and network structure, which can be used for component identification, knowledge graph formation, and theoretical improvements in learning ground truth networks.
We present a mechanism to compute a sketch (succinct summary) of how a complex modular deep network processes its inputs. The sketch summarizes essential information about the inputs and outputs of the network and can be used to quickly identify key components and summary statistics of the inputs. Furthermore, the sketch is recursive and can be unrolled to identify sub-components of these components and so forth, capturing a potentially complicated DAG structure. These sketches erase gracefully; even if we erase a fraction of the sketch at random, the remainder still retains the `high-weight' information present in the original sketch. The sketches can also be organized in a repository to implicitly form a `knowledge graph'; it is possible to quickly retrieve sketches in the repository that are related to a sketch of interest; arranged in this fashion, the sketches can also be used to learn emerging concepts by looking for new clusters in sketch space. Finally, in the scenario where we want to learn a ground truth deep network, we show that augmenting input/output pairs with these sketches can theoretically make it easier to do so.