HCC+: Hyperbolic Guarding for Certified Attention Retrieval

arXiv:2608.24971v1 Announce Type: cross Abstract: We study the Lipschitz stability of attention retrieval in hyperbolic spaces. Existing methods lack deterministic guarantees on attention-weight preservation under finite-precision representations. We introduce HCC+, a theoretical framework exploiting three properties of the Poincare ball: exponential volume growth enabling query-independent boundary truncation; logarithmic covering radius of hyperbolic 1-centers enabling dimension-independent critical-key identification; and a packing bound with constants independent of the embedding dimension. We prove two deterministic guarantees: for exact retrieval, the per-layer attention deviation is bounded by 10% of its ideal value; for soft attention, the total variation distance decays as O(1/sqrt{n}), the rate of finite-sample variance. As a consequence of the guarding mechanism, the framework achieves a storage reduction factor of 6.1imes relative to FP16. We provide the first deterministic, query-independent retrieval certificate in non-Euclidean geometry.

Related

• 1-Lipschitz Neural Networks on Hadamard Manifolds
• Not All Retrievals are Useful: Cross-Attention for Input-Aware RAG in Time Series Forecasting
• Lipschitz Dueling Bandits over Continuous Action Spaces

Source: arXiv cs.LG | 2026-08-27