This month returns with an Agentic theme, but with the added element of Self-Improvement!
FOOD UPDATE : Google has told us that there WILL BE food at the event - and it'll be "in-house" rather than pizza delivery! So: No need to eat beforehand, as far as we know.
Talks:
"RAG, Agents, RL" - Vivek Kalyan
Vivek will talk about his practical experiences and key insights training a multi turn agent using Reinforcement Learning for retrieval on legal data, achieving comparable performance to state-of-the-art models. Throughout his career, Vivek has been designing, building and evaluating machine learning products that help people make sense of data. As the cofounder of Cartograph, he developed LLM agents that automatically generate technical documentation from codebases.
"Self-Improving Agents" - Martin Andrews
In his talk, Martin will explore the new work from Sakana.ai : "Darwin Gödel Machines". Despite its intimidating title, this work is fundamentally about building an agentic system that achieves excellent results : With the additional aspect of the system being able to improve the agents through self-evaluation. Martin will also touch on two other systems : DSPy (revisited) and "GPU Kernel Scientist".
---
Talks will start at 7:15pm and end at around 8:45pm, at which point people normally come up to the front for a bit of a chat with each other, and the speakers.
As always, we're actively looking for more speakers - both '30 minutes long-form', and lightning talks. For the lightning talks, we welcome folks to come and talk about something cool they've done with keras, PyTorch, JAX and/or Deep Learning for 5-10mins (so, if you have slides, then #max=10). We believe that the key ingredient for the success of a Lightning Talk is simply the cool/interesting factor. It doesn't matter whether you're an expert or an enthusiastic beginner: Given the responses we have had to previous talks, we're sure there are lots of people who would love to hear what you've been playing with. If you're interested in talking, please just introduce yourself to Martin at one of the events.
