Symmetry breaking and nonlocality phase transitions occur nearly simultaneously during diffusion model generation in modern transformers.
hub
An analytic theory of creativity in convolutional diffusion models
11 Pith papers cite this work. Polarity classification is still indexing.
11
Pith papers citing it
hub tools
verdicts
UNVERDICTED 11representative citing papers
A minimal embedding model shows representation collapse arises from frustrated samples through slow dynamics and is prevented by stop-gradient.
A nonequilibrium latent-variable Markov model spontaneously develops cycles during likelihood training that enhance generative performance over equilibrium approaches.
Presents SimA, a score-based single-query membership inference attack for diffusion models and LDMs that uses denoiser output norm to reveal training set proximity and outperforms multi-query baselines on eight datasets.
Analytic solution of full-batch gradient flow for linear and convolutional denoisers in diffusion models yields a universal inverse-variance spectral law for learning times of eigenmodes.
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核
Flow-Direct constructs a reusable non-parametric guidance field from the log-density ratio of base and target distributions using all accumulated reward samples for feedback-efficient guidance in flow models.
Uniform-based discrete diffusion models behave as associative memories that retrieve unseen data, with a dataset-size-driven memorization-to-generalization transition detectable via conditional entropy of token predictions.
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
Diffusion models overfit denoising loss at intermediate noise but generalize in inference as model error smooths the flow field and sampling paths avoid memorized noisy training data.
The paper introduces a phase framework for data distributions connected by local denoisers and demonstrates that reverse diffusion consists of trivial and data phases separated by a transition where local score functions must fail, tied to spatial Markovianity.
citing papers explorer
-
Concurrence of Symmetry Breaking and Nonlocality Phase Transitions in Diffusion Models
Symmetry breaking and nonlocality phase transitions occur nearly simultaneously during diffusion model generation in modern transformers.
-
A Minimal Model of Representation Collapse: Frustration, Stop-Gradient, and Dynamics
A minimal embedding model shows representation collapse arises from frustrated samples through slow dynamics and is prevented by stop-gradient.
-
Emergence of Nonequilibrium Latent Cycles in Unsupervised Generative Modeling
A nonequilibrium latent-variable Markov model spontaneously develops cycles during likelihood training that enhance generative performance over equilibrium approaches.
-
Score-based Membership Inference on Diffusion Models
Presents SimA, a score-based single-query membership inference attack for diffusion models and LDMs that uses denoiser output norm to reveal training set proximity and outperforms multi-query baselines on eight datasets.
-
An Analytical Theory of Spectral Bias in the Learning Dynamics of Diffusion Models
Analytic solution of full-batch gradient flow for linear and convolutional denoisers in diffusion models yields a universal inverse-variance spectral law for learning times of eigenmodes.
-
Mechanisms of Misgeneralization in Physical Sequence Modeling
Generative sequence models for physical tasks exhibit physical misgeneralization where local prediction errors propagate through physical measurements to distort aggregate distributions over quantities like distance or energy; a data deviation kernel explains and predicts the shifts and supports a内核
-
Flow-Direct: Feedback-Efficient and Reusable Guidance for Flow Models via Non-Parametric Guidance Field
Flow-Direct constructs a reusable non-parametric guidance field from the log-density ratio of base and target distributions using all accumulated reward samples for feedback-efficient guidance in flow models.
-
Language Diffusion Models are Associative Memories Capable of Retrieving Unseen Data
Uniform-based discrete diffusion models behave as associative memories that retrieve unseen data, with a dataset-size-driven memorization-to-generalization transition detectable via conditional entropy of token predictions.
-
Generalization in LLM Problem Solving: The Case of the Shortest Path
LLMs show strong spatial generalization to unseen maps in shortest-path tasks but fail length scaling due to recursive instability, with data coverage setting hard limits.
-
Diffusion Models Memorize in Training -- and Generalize in Inference
Diffusion models overfit denoising loss at intermediate noise but generalize in inference as model error smooths the flow field and sampling paths avoid memorized noisy training data.
-
Local Diffusion Models and Phases of Data Distributions
The paper introduces a phase framework for data distributions connected by local denoisers and demonstrates that reverse diffusion consists of trivial and data phases separated by a transition where local score functions must fail, tied to spatial Markovianity.