A single math framework predicts the best way to generate images and text
Researchers show that diffusion, autoregressive, and hybrid models can all be ranked before they even start generating.

A math method that ranks how AI models should generate content before they start creating anything.
- Check the data structure and count its treedepth
- Use model weights to measure how steps depend on each other
- Rank all possible generation schedules by their hidden cost
- Pick the best schedule before any output is generated
One lattice to rule them all
Diffusion models and autoregressive (AR) models have long been treated as separate families of generative AI. Diffusion models, which gradually strip noise from an entire image or sequence at once, became the standard for continuous data like pictures and videos. Autoregressive models, which generate one token at a time conditioned on what came before, dominated discrete domains like language 1. Recent hybrid approaches have tried to blend the two, but each one locks in its own decoding schedule by design.
A new paper argues that this division is unnecessary. The researchers, T. Y. Tsui, Jiatao Gu, and Lingjie Liu from the University of Pennsylvania, present a single theoretical framework called a corruption lattice that places diffusion, AR, and every intermediate schedule on the same mathematical structure. Every possible decoder becomes a path through this lattice 1.
Cost is discarded dependence
The key insight is how they define a schedule's quality. They measure the cost of any decoding schedule as the total correlation — the dependence — that its parallel steps throw away. This cost equals the schedule's divergence from exact sampling, meaning it quantifies exactly how far a given approach strays from perfect generation 1.
From this definition emerges a clean prediction: the minimum number of steps a zero-cost schedule needs is determined entirely by the geometry of the data itself. For data that follows a Markov structure on a graph, that minimum equals the graph's treedepth — a well-known measure from graph theory. Treedepth grows logarithmically with sequence length and linearly with the side length of a grid 1. This is the same rule for both discrete tokens and continuous fields.
Predicting before decoding
When a schedule uses fewer steps than the treedepth, it must pay a positive cost — some dependence is always lost. The paper shows that you can still rank those suboptimal schedules before any decoding begins, using a pairwise-dependence kernel estimated directly from pretrained model weights 1. This means you do not need to run the full generation to know which schedule will perform best at a fixed step count.
The authors test these predictions across text generation, image generation, and video generation, verifying most of the ranking predictions under different metrics and benchmarks 1. The work gives model designers a concrete principle for choosing and constructing decoding schedules, whether they are building future AR models, diffusion models, or hybrids somewhere in between.
Code for the project has been released publicly.
Why it matters
For years, picking a decoding schedule has been largely an engineering guess — tune it empirically for each model family. This framework turns that process into something predictable. If you know the dependence structure of your data and have pretrained weights, you can rank schedules ahead of time rather than running dozens of trial generations 1. That shift from trial-and-error to theory-driven design could save significant compute across the next generation of generative models.
Key facts
- Diffusion, AR, and hybrid models can all be represented as paths on a single corruption lattice
- The minimum zero-cost steps for Markov data equal the graph's treedepth — logarithmic in sequence length
- Schedule rankings can be predicted before decoding using a pairwise-dependence kernel from pretrained weights
- Predictions were verified across text, image, and video generation benchmarks
- Code is available at https://github.com/TSUITUENYUE/The-Lattice-of-Transition-Laws
Sources
If this story concerns you or your company, or uses something that belongs to you, ask for a correction or for it to be taken down. How requests work
Keep reading
How it worksHow Meta boosts engineering productivity by 3× using agents to propagate ad model techniquesHerald of AI
How Meta boosts engineering productivity by 3× using agents to propagate ad model techniques
Meta's new AI agent system adapts small improvements across thousands of ad models automatically, saving millions in computing costs.
How it worksMIT researcher wants AI to explain itself instead of hiding how it worksHerald of AI
MIT researcher wants AI to explain itself instead of hiding how it works
Jacob Andreas is training AI to talk about what it learned — giving everyday users a way to understand how these systems make decisions, not just relying on complex debugging tools.
How it worksYibo Liu's research shows AI agents lose knowledge every time they pass notes to successorsHerald of AI
Yibo Liu's research shows AI agents lose knowledge every time they pass notes to successors
When AI systems summarize their work for the next round, some lessons vanish — a problem for any tool that relies on step-by-step memory.