Poster · Paris

GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling

Name: GSQ: Highly-Accurate Low-Precision Scalar Quantization for LLMs via Gumbel-Softmax Sampling
Start: 2026-12-09T17:00:00+01:00
End: 2026-12-09T19:00:00+01:00
Location: Paris Poster Hall, Paris

When: Wed 9 Dec 2026, 17:00–19:00 CET
Where: Paris Poster Hall
Session: Paris Poster Session 2
Authors: Alireza Dadgarnia, Rush Tabesh, Mahdi Nikdan, Michael Helcig, Eldar Kurtić, Maximilian Kleinegger, Dan Alistarh
Organisations: ISTAustria, Institute of Science and Technology Austria, ETH Zurich, NeuralMagic
QRS classification: Language models, Efficient and distributed ML systems

Open in schedule Official page