paper-with-me

홈 › Papers

On convex decision regions in deep network representations

2023-05-26 · Lenka Tětková, Thea Brüsch, Teresa Karen Scheidt, Fabian Martin Mager, Rasmus Ørtoft Aagaard, Jonathan Foldager, Tommy Sonne Alstrøm, Lars Kai Hansen

Current work on human-machine alignment aims at understanding machine-learned latent spaces and their correspondence to human representations. G{\"a}rdenfors' conceptual spaces is a prominent framework for understanding human representations. Convexity of object regions in conceptual spaces is argued to promote generalizability, few-shot learning, and interpersonal alignment. Based on these insights, we investigate the notion of convexity of concept regions in machine-learned latent spaces. We develop a set of tools for measuring convexity in sampled data and evaluate emergent convexity in layered representations of state-of-the-art deep networks. We show that convexity is robust to basic re-parametrization and, hence, meaningful as a quality of machine-learned latent spaces. We find that approximate convexity is pervasive in neural representations in multiple application domains, including models of images, audio, human activity, text, and medical images. Generally, we observe that fine-tuning increases the convexity of label regions. We find evidence that pretraining convexity of class label regions predicts subsequent fine-tuning performance.

📄 PDF Abstract BibTeX arXiv:2305.17154

Code (1)

lenkatetkova/convexity-of-representations 공식 구현 pytorch

Tasks

Few-Shot Learning

Similar Papers 제목 키워드 기반

Can Neural Nets Learn the Same Model Twice? Investigating Reproducibility and Double Descent from the Decision Boundary Perspective

2022-03-15 · CVPR 2022 1 · Gowthami Somepalli, Liam Fowl, Arpit Bansal, Ping Yeh-Chiang 외

We discuss methods for visualizing neural network decision boundaries and decision regions. We use these visualizations to investigate issues related to reproducibility and generalization in neural network training. We o…

Inductive Bias

Decision-Focused Federated Learning Under Heterogeneous Objectives and Constraints

2026-04-21 · Konstantinos Ziliaskopoulos, Alexander Vinel arxiv

We consider Decision-Focused Federated Learning (DFFL), a predict-then-optimize setting in which multiple clients collaboratively train predictive models for downstream linear optimization problems without exchanging raw…

Federated Learning

Entity Embeddings with Conceptual Subspaces as a Basis for Plausible Reasoning

2016-02-18 · Shoaib Jameel, Steven Schockaert

Conceptual spaces are geometric representations of conceptual knowledge, in which entities correspond to points, natural properties correspond to convex regions, and the dimensions of the space correspond to salient feat…

Entity Embeddings

Convex Collision-Free Regions

2026-07-29 · Tomoyo Kikuchi, Takashi Kanai arxiv

Convex Collision-Free Regions (CCFR) is a collision handling method that explicitly represents local convex feasible regions to enforce non-penetration. Each feasible region is constructed from surrounding mesh primitive…

Partition-wise Linear Models

2014-10-31 · NeurIPS 2014 12 · Hidekazu Oiwa, Ryohei Fujimaki

Region-specific linear models are widely used in practical applications because of their non-linear but highly interpretable model representations. One of the key challenges in their use is non-convexity in simultaneous …