Random rotation
A random rotation is an orthogonal transformation selected randomly to change the orientation of vectors while preserving their lengths and angles.
Quantization and Fast Inference
A random rotation is an orthogonal transformation selected randomly to change the coordinate orientation of numerical vectors.
- How it is used
- Redistributes normalized KV-cache vector coordinates before scalar quantization in TurboQuant.
- How deep it goes
- Explains orthogonality, the resulting coordinate distribution, and the quantization role, then shows matrix construction in PyTorch.
- Examples
- Uses a 128-dimensional attention-head vector and a card-shuffling analogy to illustrate coordinate redistribution.
- What you can do after
- Apply it by generating an orthogonal matrix with seeded PyTorch QR decomposition and rotating normalized vectors.
Read how Quantization and Fast Inference explains random rotation →
Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI
A random rotation is a reversible transformation that applies a randomly selected orthogonal rotation to input feature vectors.
- How it is used
- Uses reversible feature transformations to compare model behavior and expose how orientation affects tree-based learning.
- How deep it goes
- Defines the matrix properties and explains model trade-offs, but does not provide implementation steps or derivations.
- Examples
- Uses rotated tabular data to contrast gradient-boosted trees with deep neural networks.
- What you can do after
- Apply a reversible rotation before modeling to test orientation sensitivity and compare tree-based with deep-learning approaches.
Related concepts
- TurboQuantused by · TurboQuant is a data-oblivious vector quantization method that separates vector magnitude from direction and applies random rotation before low-bit quantization.
- gradient boostingcontrasts with · Gradient boosting is an ensemble-learning technique that sequentially adds predictive models fitted to reduce a loss by following its gradient.
Questions about random rotation
What is random rotation?
A random rotation is an orthogonal transformation selected randomly to change the orientation of vectors while preserving their lengths and angles.
Which Manning books explain random rotation?
Quantization and Fast Inference and Machine Learning for Tabular Data: XGBoost, Deep Learning, and AI.
What is the difference between random rotation and gradient boosting?
A random rotation is an orthogonal transformation selected randomly to change the orientation of vectors while preserving their lengths and angles. Gradient boosting is an ensemble-learning technique that sequentially adds predictive models fitted to reduce a loss by following its gradient.
These guides are generated from the text of Manning books by Manning's knowledge graph and link to the sections they draw on. The full books are available on liveBook and at manning.com.

