Index

Metric for Cognitive Interest

Resources paper

Top Advantages

• Quantitative metric to measure the 'interestingness' of text.

New Things

• A metric for measuring interestingness of text. Measures predictive inference caused by foreshadowing in text.
• Uses word vectors and cosine similiarities in an ionteresting way

Model Summary

Notable Insights Using 1x1 conv to speed up convolutions. Check [here](https://stats.stackexchange.com/questions/194142/what-does-1x1-convolution-mean-in-a-neural-network#:~:text=A 1x1 convolution simply maps,volumes with extremely large depths.)

Misc things

• Very good related work/intro section. Nicely covers some cognitive science concepts for texts.

limitations

• I didn't find the results persuasive enough. Plus the survey to test results were very limited in terms of quantity.

Questions

Tags metric, perception, cognitive

Papers To Add Here

- PCGML review paper
- Mario as text seq