Freeze, Diffuse, Decode (FDD), a novel diffusion-based framework that adapts pre-trained embeddings to downstream tasks while preserving their underlying geometric structure, is introduced.
Abstract
Pretrained transformers provide rich, general-purpose embeddings, which are transferred to downstream tasks. However, current transfer strategies: fine-tuning and probing, either distort the pretrained geometric structure of the embeddings or lack sufficient expressivity to capture task-relevant signals. These issues become even more pronounced when supervised data are scarce. Here, we introduce Freeze, Diffuse, Decode (FDD), a novel diffusion-based framework that adapts pre-trained embeddings to downstream tasks while preserving their underlying geometric structure. FDD propagates supervised signal along the intrinsic manifold of frozen embeddings, enabling a geometry-aware adaptation of the embedding space. Applied to antimicrobial peptide design, FDD yields low-dimensional, predictive, and interpretable representations that support property prediction, retrieval, and latent-space interpolation.
This work argues that cross-modal alignment is implicitly captured in the information-compression trajectory, and proposes LLaVAFlow, an information-theoretic distillation framework that preserves alignment flow and enhances both downstream performance and generalization.
Mu-Yao Yuan, Mu-Yan Jiao, Jiang-Yong Ying et al.· 0 citations
This work proposes Explainable Probing of Cross-Domain Sparse Embeddings (EXPOSE), a framework that uses Sparse Autoencoders (SAEs) as an explainable bottleneck to identify and suppress domain-specific components in VFM embeddings.
Anja Witte, M. Lennartz, Jan Baumbach et al.· 0 citations
Aether is introduced, a simple plug-in method that applies diffusion-style random perturbations in the embedding space via controlled alpha-mixing, specifically designed to provide isotropic regularization that remains semantically consistent.
Hyesong Choi, Daeun Kim, Song Park et al.· 0 citations
This work introduces REPI, a training framework based on a scaffold-to-internalization strategy, in which projected encoder representations initially serve as a temporary scaffold and are then progressively internalized by the diffusion transformer.
Han-Si Fu, Jia-Cheng Chen, Bao-Quan Zhao et al.· 0 citations
Predicting macroscopic properties of crystalline materials from atomic structure remains a central challenge in computational materials science. We introduce Pformer, a modular crystal representation framework built around a Fourier-inspired source encoder. The source model is trained by supervised scalar-property regr...
Position-prompted PCA (P3CA), an encoder-agnostic method for local probing of channel-rich spatial tensors, is proposed and implemented in EmbedVision, an interactive 3D Slicer-based workflow, and evaluated across natural images, colorectal pathology foundation-model embeddings, and spatial transcriptomic tensors.
A. Jamzad, Dilakshan Srikanthan, F. Akbarifar et al.· 0 citations
Learn how Google DeepMind and GovTech Singapore uses lightweight on-device Gemma models to intercept conversational scams in private chats with zero cloud data egress.
MIT News · Artificial Intelligence· news.mit.eduOct 13, 2026
Machine-learning networks can learn without detailed instructions from a designer. Those that do may offer insight into how the brain reorganizes itself during learning.
We use cookies to run the site and, with your consent, for analytics and to show ads.
See our Cookie Policy.
