Improving Generative Model Self-Training with Geometrically Modified Outputs

arXiv:2609.35512 · cs.LG, cs.AI · Submitted 2026-09-28 · Read on arXiv

cs.LG, cs.AI

Submitted: 2026-09-28

Updated: 2026-09-28

Code: https://github.com/lumalabs/imm4https:

Terminology

• Generative Model — 1× in this paper · explained in 2 episodes

Sources

• Self-Improving Diffusion Models with Synthetic Data
• Generative Modeling via Drifting
• Scaling Laws for Autoregressive Generative Modeling
• Scaling Laws for Neural Language Models
• Refining Generative Process with Discriminator Guidance in Score-based Diffusion Models
• Scaling Data-Constrained Language Models
• Consistency Models
• Representation Fr'echet Loss for Visual Generation
• Image Generation with a Sphere Encoder

Related papers

• Polynomial-Augmented Neural Networks (PANNs) with Weak Orthogonality Constraints for Enhanced Function and PDE Approximation
• AIRL-S: Unifying Reinforcement Learning and Search-Based Test-Time Scaling via Adversarial Inverse Reinforcement Learning
• Transformers as Bayesian In-Context Experimenters: Smoothness-Adaptive Efficient ATE Estimation
• Convergence issues in Relational Concept Analysis based on AOC-posets
• Beliefs Beyond Posteriors: The Predictive Cost of ELBO Binding
• Understanding Diffusion Models via Ratio-Based Function Approximation with SignReLU Networks