Journal Club, episode 4

Lan presents a blog post from Google Deepmind about Dopamine and temporal difference learning. This is the story of a fruitful collaboration between Neuroscience and AI researchers that found the activity of dopamine neurons in the mouse ventral tegmental area during a learnt probablistic reward task was consistent with distributional temporal-difference reinforcement learning. That’s a mouthful, go read it yourself!

Kyle brings the news item Hey Alexa! Sorry I fooled you … from MIT News informing listeners that the adversarial attacks are possible on natural language, not just computer vision.

George talks about his first attempts at designing an Auto-Trading Agent with Deep Q Networks and how he was outsmarted by a group of 20 neurons.