Token rankings from language models are unique and NP-hard to forge, providing the first polynomially unforgeable model signature.
and Turk-Browne, Nicholas B
4 Pith papers cite this work. Polarity classification is still indexing.
4
Pith papers citing it
years
2026 4representative citing papers
MemPalace's high retrieval performance comes from verbatim storage and default vector database techniques, not its spatial metaphor, though it contributes low wake-up costs and deterministic offline operation.
Quasi-F^e-splitting for all e implies numerically log canonical for numerically Q-Gorenstein normal singularities, with converse in dim 2 when p does not divide the Gorenstein index, plus a classification of 2D quasi-F-split cases.
EVAF and test-retest protocol show selective parametric consolidation of high-valence experiences in GPT-2 and TinyLlama while preserving factual retrieval.
citing papers explorer
-
Token Rankings are Unforgeable Language Model Signatures
Token rankings from language models are unique and NP-hard to forge, providing the first polynomially unforgeable model signature.
-
Spatial Metaphors for LLM Memory: A Critical Analysis of the MemPalace Architecture
MemPalace's high retrieval performance comes from verbatim storage and default vector database techniques, not its spatial metaphor, though it contributes low wake-up costs and deterministic offline operation.
-
Quasi-$F$-splitting versus log canonicity
Quasi-F^e-splitting for all e implies numerically log canonical for numerically Q-Gorenstein normal singularities, with converse in dim 2 when p does not divide the Gorenstein index, plus a classification of 2D quasi-F-split cases.
-
EVAF: A Test-Retest Protocol for Selective Parametric Consolidation
EVAF and test-retest protocol show selective parametric consolidation of high-valence experiences in GPT-2 and TinyLlama while preserving factual retrieval.