Tag
This paper investigates the impact of using hyperbolic versus Euclidean latent geometry for tree-structured regularizers in prototype networks, showing that hyperbolic prototypes significantly improve local topology preservation in hierarchical classification on WikiArt.
The paper proposes FlatLand, a personalized federated learning method that uses tailored Lorentz space in hyperbolic geometry to handle heterogeneous graph structures among clients, improving performance in privacy-preserving collaborative training.
HyperSAE is an open-source Python library that uses hyperbolic geometry to organize LLM learned concepts into browsable tree structures, improving on flat feature lists. It captures 99.8% of Gemma-2-2B's features and includes interactive demos.
This paper shows that small hyperbolic language models can exhibit creativity, honesty, and designed forgetting, and introduces a behavioral auditor that detects compliance gaps humans and frontier judges miss, offering a route to trustworthy companion AI.
This paper proposes Equivariant Poincaré ResNets, combining hyperbolic geometry with discrete symmetry groups to improve efficiency in learning visual representations by treating rotated features as symmetric rather than distinct hierarchical concepts.
A solo-built multi-agent cognitive architecture uses hyperbolic geometry on a Poincaré ball manifold, variational free energy for belief updating, and wave interference for memory retrieval, allowing personality to emerge from memory interactions rather than scripting.
GoCoMA is a multimodal framework using hyperbolic Poincaré ball embeddings to fuse code stylometry and binary artifact images for attributing LLM-generated code, outperforming unimodal and Euclidean baselines on two benchmarks.
This paper introduces HSG (Hyperbolic Scene Graph), a scene graph model that leverages hyperbolic geometry for representing hierarchical scene structures. It is hosted on Hugging Face and referenced via arXiv:2604.17454.