Is the ICLR 2026 review deadline really Oct 31? 5 papers in 15 days, that’s just begging for bad reviews 🤦‍♂️ 🤦‍♂️

Every time I see some SgLang or VLLM only feature being talked about

I’ll be at #NeurIPS2025 from 2nd-6th Dec. DM if you want to chat about MoEs, scaling training and inference, making GPUs go brrr and other fun stuff !!

Not going back to calling it GPT ever again

Will be at #NeurIPS2025 from 2nd to 6th Dec. Excited to chat about async RL, Environment Exploration, Agents/Tool use, User Simulator, Synthetic Data Generation or any other topic!! You can find me at the Databricks booth @ Tue 12 - 4pm

Riding this wave for a second Chris Barber! Thank you for putting together an amazing list!!! Databricks Mosaic Research is also hiring for researchers and engineers for many projects / at all levels. DM me or come hang out with the team at our social: luma.com/regcpdhi

look how they massacred my boy 😭

Price of brick go up !!!! 🧱🧱

Come work on challenging problems with a team pushing boundaries for enterprise agentic capabilities !!!

Crisp writing, worth reading for everyone building today !!

This was not in my bingo card for 2025!! More thoughts later on but huge congrats to Jonathan Ross and co!!

Came up for air to find this Christmas 🎄 gift !!! Skyler Miao and team killing it

Reliable enterprise agents require system-level reasoning when retrieving across heterogeneous knowledge sources. Traditional RAG often fails to consistently follow instructions, schemas, and constraints end to end. That’s why we’re presenting Instructed Retriever, a new

Instructed Retriever is a multi-tiered declarative approach for building high quality search agents. It's an example of an "instructed system", which goes beyond prompt tuning and tool calling by passing data among modules which work together to fulfill an information need.

Is this the same dude from 2021? 🙄

OpenAI and @Cerebras have signed a multi-year agreement to deploy 750 megawatts of Cerebras wafer-scale systems to serve OpenAI customers. This has been a decade in the making. Deployment begins in early 2026, and when fully rolled out, it will be the largest high-speed AI

Agent memory is a simple and powerful way to do continual learning! With the new MemAlign method from Databricks Research, we can build better LLM judges from examples of human ratings, and they scale with more data. Now in Databricks and MLflow. databricks.com/blog/memalign-…

I now constantly get questions about the SAAS meltdown, role of AI, system of records etc. I don't have an answer to all these. But I do know that we saw an acceleration in our business in Q2, Q3, and now finished the year with accelerating Q4. The question is, why? Short

when you see too many NCCL timeouts 🥲🥲