Twitter FastText and Word2vec embeddings.
In my PhD thesis, I made a comparison between FastText and Word2vec word embeddings for PoS tagging and NER in Twitter microposts. You can find a blog post over here and you can find the embeddings on Github.
Contextual Decomposition for CNNs
As part of my paper, I have implemented a decomposition technique for convolutional neural networks. I allows you to understand which patterns the neural networks is looking for to make a certain decision. In my case, I was interested in comparing character-level patterns of CNNs and BiLSTMs. The implementation can be found here.
Twitter Word2vec model (WNUT Challenge)
As part of our ACL W-NUT 2015 shared task paper, we release a Twitter word2vec model trained on 400 million tweets, as described in detail in this paper. The code for this is currently not available but I recommend using the newer embeddings which you can find on Github.