Review: Generative Models, Multimodal Learning, and Closed-Loop Workflows in Inverse Materials Design
This arxiv review surveys recent advances in generative modeling for inverse materials design, covering variational autoencoders, normalizing flows, autoregressive models, and diffusion models applied to crystalline solid discovery. It examines how multimodal learning fuses crystal structures, thermodynamic data, spectroscopy, microscopy, and scientific text into transferable chemical-space representations. The paper also reviews closed-loop design pipelines integrating conditional generation with Bayesian optimization, reinforcement learning, and active learning, and identifies recurring failure modes including surrogate exploitation, diversity collapse, and the stability-synthesizability gap.
Related guides (3)
Related events (8)
Glow: Better reversible generative models
OpenAI introduces Glow, a reversible generative model using invertible 1x1 convolutions that extends prior work on normalizing flows. The model generates realistic high-resolution images, supports efficient sampling, and learns disentangled features for attribute manipulation. Code and an online visualization tool are released alongside the paper.
OpenAI: Generative Models Overview (2016)
A 2016 OpenAI blog post describing four research projects centered on generative models as a branch of unsupervised learning. The post explains what generative models are, their importance, and potential future directions. This is an archival piece predating modern large language models and diffusion systems, representing early foundational work at OpenAI.
PLAID: Repurposing Protein Folding Models for Multimodal Protein Generation with Latent Diffusion
PLAID is a generative model that simultaneously produces protein 1D sequences and 3D all-atom structures by learning a diffusion model over the latent space of ESMFold, a protein folding model. It requires only sequence data for training—leveraging databases 2-4 orders of magnitude larger than structure databases—and decodes structure at inference via frozen folding model weights. The approach supports compositional prompting by function and organism, addressing practical drug-design constraints like humanization and solubility. A companion compression model, CHEAP, addresses the high-dimensionality of transformer latent spaces to make the diffusion training tractable.
Joint Energy-Based Models Reveal a Generative-Discriminative Sweet Spot for Human-Aligned Vision
Researchers use Joint Energy-Based Models (JEMs) to isolate the effect of learning objective—independent of architecture, scale, and data—on human alignment in visual representations. By varying a single mixing coefficient between discriminative and generative training, they evaluate models across six human-alignment benchmarks and find that alignment peaks at intermediate points on the generative-discriminative continuum rather than at either extreme. The results suggest that hybrid objectives combining categorical structure from discriminative learning with input-structure sensitivity from generative learning yield the most human-like visual behavior. This challenges the framing of generative vs. discriminative as a binary choice for building human-aligned vision systems.
Radical AI's Joseph Krause argues the lab infrastructure is the moat in AI-driven materials science
Latent Space interviews Joseph Krause of Radical AI about their 'self-driving lab' approach to materials discovery, where automated physical experimentation is the core differentiator rather than the underlying AI model. Krause argues that in materials science, the data generation pipeline and lab automation create defensible advantages that model capabilities alone cannot replicate. The piece highlights a deployment pattern where AI is tightly coupled with physical-world feedback loops in scientific research.
The Annotated Diffusion Model
A Hugging Face blog post providing a detailed, annotated walkthrough of diffusion models for image generation, likely covering the mathematical foundations and implementation details of denoising diffusion probabilistic models (DDPMs). The post serves as an educational deep-dive into the architecture and training process of diffusion-based generative models. Published in mid-2022, it coincides with the period of rapid growth in diffusion model adoption.
Survey: Agentic Environment Engineering for LLMs — Modeling, Synthesis, Evaluation, and Application
A comprehensive arXiv survey systematically reviews the design and engineering of interactive environments for LLM-based agents, covering the full lifecycle from environment modeling and synthesis to evaluation and application. The paper categorizes environments across eight attributes and eight domains, introduces symbolic and neural synthesis paradigms, and characterizes four pathways for agent-environment co-evolution including memory-centric, orchestration-centric, trajectory-centric, and exploration-centric approaches. It also identifies three paradigms of environment evolution (neural-driven, difficulty-driven, scaling-driven) and proposes future directions such as Environment-as-a-Service and multi-agent environments. This is a reference-organizing contribution for the rapidly growing agent tooling and evaluation space.
Training Design for Text-to-Image Models: Lessons from Ablations
Photoroom shares practical lessons from ablation studies on training design choices for text-to-image diffusion models. The post covers decisions around data curation, model architecture, and training hyperparameters derived from systematic experimentation. This is part two of a series documenting Photoroom's internal research into building production-grade image generation systems.