QuIP#: Even Better LLM Quantization with Hadamard Incoherence and Lattice Codebooks

2 code implementations • 6 Feb 2024 • Albert Tseng, Jerry Chee, Qingyao Sun, Volodymyr Kuleshov, Christopher De Sa

Post-training quantization (PTQ) reduces the memory footprint of LLMs by quantizing their weights to low-precision.

Quantization

Beyond Invariance: Test-Time Label-Shift Adaptation for Distributions with "Spurious" Correlations

1 code implementation • 28 Nov 2022 • Qingyao Sun, Kevin Murphy, Sayna Ebrahimi, Alexander D'Amour

However, we assume that the generative model for features $p(x|y, z)$ is invariant across domains.

Individualized and Global Feature Attributions for Gradient Boosted Trees in the Presence of $\ell_2$ Regularization

1 code implementation • 8 Nov 2022 • Qingyao Sun

We propose Prediction Decomposition Attribution (PreDecomp), a novel individualized feature attribution for gradient boosted trees when they are trained with $\ell_2$ regularization.

Additive models feature selection

Action2Motion: Conditioned Generation of 3D Human Motions

1 code implementation • 30 Jul 2020 • Chuan Guo, Xinxin Zuo, Sen Wang, Shihao Zou, Qingyao Sun, Annan Deng, Minglun Gong, Li Cheng

Action recognition is a relatively established task, where givenan input sequence of human motion, the goal is to predict its ac-tion category.

Action Generation