Erik Lindgren

[Erik Lindgren @erikmlindgren]

Super interesting and surprising results from @mat_jacob1002 and team on reranker quality. Check it out!

[Mathew Jacob @mat_jacob1002]

It's time to revisit common assumptions in IR! Embeddings have improved drastically, but mainstream IR evals have stagnated since MSMARCO and BEIR.

We ask: on private or tricky IR tasks, are current rerankers even better? Surely, reranking as many docs as you can afford is best?

[Jonathan Frankle @jefrankle]

Meet DBRX, a new sota open llm from @databricks. It's a 132B MoE with 36B active params trained from scratch on 12T tokens. It sets a new bar on all the standard benchmarks, and - as an MoE - inference is blazingly fast. Simply put, it's the model your data has been waiting for.

[Alex Dimakis @AlexGDimakis]

Video for our NeurIPS paper Experimental Design for Cost-Aware Learning of Causal Graphs
Smoothly narrated by Erik's radio voice.
https://t.co/kl3xbwwKIs
Poster: Thursday 10:45 AM -- 12:45 PM @ Room 210 #1
Paper: https://t.co/oFuAjZ5H23