Just finished making slides for my tomorrow's talk at Lossfunk! I'll give 5 counterintuitive insights about why these things work so well. and will fit a 1.8 million parameter model on just 100 data points! Register for in-person (bangalore) or online below 👇

this is Lossfunk in one statement!

😈

Was reading the ultra-dense Continuous Thought Machines paper by Sakana AI and got Claude to say this 👇 Now I need another LLM to know whether it's been just sycophantic or am I actually right!

Update on Lossfunk research internship. 730 applications -> 150 shortlists -> kickstarted 10 projects already! Kinds of projects: • LLMs in prediction markets • World models for Atari-100k • Soft rewards (perplexity based) v/s Hard rewards • Meta-RL (learning to learn)

awesome sunday weather and view from Lossfunk

'Differential programming is not just deep learning' starting out the genius book - 'The Elements of Differential Programming' on the suggestion of Nilesh Trivedi for his seminar in Lossfunk

We're looking forward to hosting Nilesh Trivedi today. Attend his talk in person or online 👇

Work with me! I'm expanding my team of executive assistants to help make Lossfunk the coolest AI lab out there. Reshare for reach 🙏 [Apply in the form below]

New bio :)

Nilesh Trivedi's talk at Lossfunk today on Differentiable Programming was amazing. We need more such talks. Look forward to the slides and Youtube video.