4 papers
Fast Recovery for LLM Serving via Decoupled Device Memory Lifetime in Dynamo
Schwinn Saereesitthipitak, Mohammed Abdulwahhab, Hannah Zhang +6
Large language model (LLM) inference replicas run across tightly coupled GPUs and serve traffic continuously for weeks. Hardware and software failures are therefore inevitable, and…
Nemotron 3 Ultra: Open, Efficient Mixture-of-Experts Hybrid Mamba-Transformer Model for Agentic Reasoning
NVIDIA, :, Aaron Blakeman +571
We introduce Nemotron 3 Ultra, a 550 billion total and 55 billion active parameter Mixture-of-Experts Hybrid Mamba-Attention language model. We pre-trained Nemotron 3 Ultra on 20 t…
The tree property on long intervals of regular cardinals
James Cummings, Yair Hayut, Menachem Magidor +3
In this paper we prove that the tree property can hold on regular cardinals in an interval which overlaps a strong limit cardinal. This is a crucial milestone in the long term proj…
Unreachability of Inductive-Like Pointclasses in
Derek Levinson, Itay Neeman, Grigor Sargsyan
Hjorth proved from that there is no sequence of distinct sets of length . Sargsyan extended Hjorth's technique to show there is no sequence of distinc…