Lossfunk
@lossfunk
the AI lab for independent researchers
Our co-founder @diwanksingh is speaking at @lossfunk tomorrow on Shipping Code 10X faster with Claude Code. Register below to join virtually or in-person 👇
Sneak peek into one of our research projects! Can VLMs play pong (without any scaffolding)? Not yet, but we're getting close.
New @lossfunk talk dropped! 👇 Listen to @_drewkirchhoff talk about his vision of a universal user embedding, where you ingest what different services know about you to get a deep understanding about yourself.
This new @lossfunk article is full of some of the wackiest practical tips for context engineering. Very interesting read.
Kickstarted Batch 5 of @lossfunk residency. 🧵 of who they are + their projects
What I shared with research interns at @lossfunk on how to go about their research projects.
New @lossfunk talk dropped! Hear @navbenny talk about why AI will completely dominate financial markets going forward! No human will be able to compete when GPUs go brr..
before you read this, start playing this in the background youtu.be/8qrriKcwvlY cool? now we're good to go cracked. whacked. unstoppable. professor x built a school for mutants; a place where the extraordinary could test their limits, hone their gifts, and learn to wield…
Yep, we're building a third space for AI beyond academia and industry. @lossfunk is an experiment to discover what could organizations of the future look like!
Puzzled by what your RL agent is doing? Catch my talk on “Reinforcement Learning with Function Approximation — Done Right” tomorrow (09th July) at @lossfunk. RSVP → lu.ma/ig707k8x. #ReinforcementLearning
Policy Iteration’s super-power—monotonic improvement + guaranteed convergence—vanishes under general function approximation. To bring them, we introduce Reliable Policy Iteration (RPI) : arxiv.org/abs/2506.07134. #ReinforcementLearning #RL @EshwarSR @today_itself @DalalGal
had such a great time at lossfunk! was fortunate to give a talk on my recent work (will be available soon!) and have discussions on how to empower india to contribute to research in the best conferences in the world. the commitment to foundational research here at lossfunk is…
omw to visit the coolest ai lab in india 👀
Yep, we're building a third space for AI beyond academia and industry. @lossfunk is an experiment to discover what could organizations of the future look like!
science as a loop of learning about a phenomena with empirical evidence vs research as reviewing existing literature about the study of a phenomena vs engineering as building a system that either enables or captures the study of the phenomena
Our manifesto on doing good science in AI Read the full text 👇