Andrew Drozdov

Using 100+ tokens to answer 2 + 3 =

So many cool papers this week, and it's only Tuesday.

This week I'll have a long talk with @yacinemahdid.bsky.social about KARL and all things retrieval, and couldn't be looking forward to it more. substack.com/@yacinelearn...

We built a thing! The Databricks Reranker is now in Public Preview. It's as easy as changing the arguments to your vector search call, and doesn't require any additional setup. Read more: www.databricks.com/blog/reranki...

The transformer was invented in Google. RLHF was not invented in industry labs, but came to prominence in OpenAI and DeepMind. I took 5 of the most influential papers (black dots) and visualized their references. Blue dots are papers that acknowledge federal funding (DARPA, NSF).

The PhD is pretraining. Interview prep is alignment. Take this to heart. :)

We have updated #nocha, a leaderboard for reasoning over long-context narratives 📖, with some new models including #Gemini 2.5 Pro which shows massive improvements over the previous version! Congrats to #Gemini team 🪄 🧙 Check 🔗 novelchallenge.github.io for details :)

MLflow is on BlueSky! Follow @mlflow.org to keep up to date on new releases, blogs and tutorials, events, and more.

"How Claude Code is using a 50-Year-Old trick to revolutionize programming"

Somehow my most controversial take of 2025 is that agents relying on grep are a form of RAG.

I do want to see aggregate stats about the model’s generation and total reasoning tokens is perhaps the least informative one.

"All you need to build a strong reasoning model is the right data mix." The pipeline that creates the data mix:

It’s pretty obvious we’re in a local minima for pretraining. Would expect more breakthroughs in the 5-10 year range. Granted, it’s still incredibly hard and expensive to do good research in this space, despite the number of labs working on it.

Word of the day (of course) is ‘scurryfunging’, from US dialect: the frantic attempt to tidy the house just before guests arrive.

... didn't know this would be one of the hottest takes i've had ... for more on my thoughts, see drive.google.com/file/d/1sk_t...

feeling a but under the weather this week … thus an increased level of activity on social media and blog: kyunghyuncho.me/i-sensed-anx...

Smarter, Better, Faster, Longer: A Modern Bidirectional Encoder for Fast, Memory Efficient, and Long Context Finetuning and Inference Introduces ModernBERT, a bidirectional encoder advancing BERT-like models with 8K context length. 📝 arxiv.org/abs/2412.13663 👨🏽‍💻 github.com/AnswerDotAI/...

State Space Models are Strong Text Rerankers Shows Mamba-based models achieve comparable reranking performance to transformers while being more memory efficient, with Mamba-2 outperforming Mamba-1. 📝 arxiv.org/abs/2412.14354

I’m being facetious, but the truth behind the joke is that OCR correction opens up the possibility (and futility) of language much like drafting poetry. For every interpreted pattern for optimizing OCR correction, exceptions arise. So, too, with patterns in poetry.

Reasoning is fascinating but confusing. Is reasoning a task? Or is reasoning a method for generating answers, for any task?

The future of AI is models that generate graphical interfaces. Instead of the linear, low-bandwidth metaphor of conversation, models will represent themselves to us as computers: rich visuals, direct manipulation, and instant feedback. willwhitney.com/computing-in...

Three must read papers for PhD students. #scisky #PhD #science #research #academicsky 1. The importance of stupidity in scientific research Open Access journals.biologists.com/jcs/article/...

In any given year there are between one and three Friday the 13ths. This year there are two. 👻

Reflections on NeurIPS: There's always a big theme people seem to be preoccupied with. This year, it was the continuation of scaling/progress. Will it continue? What will the next generation of models hold? I even got to sass Dylan Patel (not on bsky) over it. Here are my personal thoughts 🧵

The databricks have arrived

Slides are up! I presented on "Presentation & Consumption in the context of REML" The full deck is here. There's a lot of gems if you're interested in this space! retrieval-enhanced-ml.github.io/sigir-ap2024...

Today we'll be presenting the Tutorial on Retrieval-Enhanced Machine Learning (REML). Come by to learn about the emerging design patterns in this space and see how to use retrieval beyond RAG. In collaboration w/ the amazing @841io.bsky.social @teknology.bsky.social Alireza Salemi and Hamed Zamani.

Today we'll be presenting the Tutorial on Retrieval-Enhanced Machine Learning (REML). Come by to learn about the emerging design patterns in this space and see how to use retrieval beyond RAG. In collaboration w/ the amazing @841io.bsky.social @teknology.bsky.social Alireza Salemi and Hamed Zamani.

Few things that are revolutionizing retrieval research right now: 1. content-based models have gotten much better 2. synthetic data has increased the value of small specialized datasets 3. retrieval is becoming a more important component in a variety of AI applications

Reading articles that 🦋 might get into advertising, and I'm not sure this is what they meant 😅

2024: For really hard problems, you go to that one friend down the street who has an o1 pro subscription. 1960: For really important calls, you go to that one friend down the street who was able to get a telephone.

👀

Somehow missed this thread from @sungkim.bsky.social --- thanks for the interest in our work!

I’m on the academic job market this year! I’m completing my @uwcse.bsky.social @uwnlp.bsky.social Ph.D. (2025), focusing on overcoming LLM limitations like hallucinations, by building new LMs. My Ph.D. work focuses on Retrieval-Augmented LMs to create more reliable AI systems 🧵

RAG still has a way to go. (this book doesn’t exist)

An underlooked feature of arXiv is it provides a unified interface for all the conferences.

Every individual agent is just a sparse instantiation of the mother agent that represents all agents.