HBCT lifts embeddings into Lorentz hyperbolic space, uses entailment cones to keep new embeddings inside old ones' cones, and weights contrastive alignment by an uncertainty estimate, improving backward-compatible retrieval in experiments.
Stationary representations: Optimally approximating compatibility and implications for improved model replacements
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.LG 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
Learning Along the Arrow of Time: Hyperbolic Geometry for Backward-Compatible Representation Learning
HBCT lifts embeddings into Lorentz hyperbolic space, uses entailment cones to keep new embeddings inside old ones' cones, and weights contrastive alignment by an uncertainty estimate, improving backward-compatible retrieval in experiments.