AI Distillery
Automatically modelling and distilling knowledge within AI!

Created by The M Tank

Tell Me More


Our plan, our mission and our goals.

Medium blog series detailing AI Distillery:
AI Distillery (Part 1): A bird's eye view of AI research
AI Distillery (Part 2): Distilling by Embedding

You can also read about MTank's general vision here.

Modeling AI Knowledge

Handling AI knowledge is becoming increasingly difficult due to the large amount of paper published on open access platforms like Effectively modeling data is necessary in order to allow people to access and explore information freely.

Automatically Extracting Knowledge from Data

Due to the large amount of data, knowledge extraction cannot be a manual process: we employ automatic techniques to extract information from papers and use this to analyze them.

Sharing Knowledge with the Community

We extract knowledge and we give free access to it for the community. Everybody should be able to access to this data and to use it. Even a novice with little understanding of AI should be given a place to see easily where to AI knowledge progressing.

Search and Proximity Pages

Search for papers with free-text or find similar word/paper embeddings to given words/papers

Paper Search


Word Embedding Proximity

Paper Embedding Proximity

Visualization Pages

Look at our embedding visualizations and explore arxiv data with our charts

Word Embedding Visualization

Paper Embedding Visualization

Charts and Additional Insights

The AI Distillery team

Parent organisation: The M Tank

Ben Duffy


Lukas Galke


Federico Bianchi


We have experience with Natural Language Processing (NLP), Deep Learning (DL), Information Retrieval (IR), Web Development and Visualisation