An intrinsic effective sample size for manifold MCMC is defined via kernel discrepancy as the number of independent draws yielding equivalent expected squared discrepancy to the target.
Quantization.IEEE Transactions on Information Theory, 44(6):2325–2383
7 Pith papers cite this work. Polarity classification is still indexing.
7
Pith papers citing it
representative citing papers
The profile maximum likelihood estimator for the location in anisotropic hyperbolic wrapped normal models is strongly consistent, asymptotically normal, and attains the Hájek-Le Cam minimax lower bound under squared geodesic loss.
Low-bit post-training quantization of reasoning LLMs increases reasoning token counts while preserving accuracy, introducing a hidden test-time compute cost.
NSVQ mitigates codebook collapse in large-codebook VQ by addressing encoder drift via non-stationary loss, replacement, and staged freezing, improving rFID from 2.39 to 2.10 on ImageNet-1k while achieving 100% utilization.
A pipeline using product quantization and systematic parameter evaluation creates data-driven soil taxonomies with higher specificity than human-derived classifications.
Dask parallelization of product quantization and inverted indexing allows large-scale approximate nearest neighbor search while preserving accuracy and reducing computation to medium-scale levels.
citing papers explorer
-
Intrinsic effective sample size for manifold-valued Markov chain Monte Carlo via kernel discrepancy
An intrinsic effective sample size for manifold MCMC is defined via kernel discrepancy as the number of independent draws yielding equivalent expected squared discrepancy to the target.
-
Profile Likelihood Inference for Anisotropic Hyperbolic Wrapped Normal Models on Hyperbolic Space
The profile maximum likelihood estimator for the location in anisotropic hyperbolic wrapped normal models is strongly consistent, asymptotically normal, and attains the Hájek-Le Cam minimax lower bound under squared geodesic loss.
-
Quantization Inflates Reasoning: Token Inflation as a Hidden Cost of Low-Bit Reasoning Models
Low-bit post-training quantization of reasoning LLMs increases reasoning token counts while preserving accuracy, introducing a hidden test-time compute cost.
-
NSVQ: Mitigating Codebook Collapse by Stabilizing Encoder Drift in Vector Quantization
NSVQ mitigates codebook collapse in large-codebook VQ by addressing encoder drift via non-stationary loss, replacement, and staged freezing, improving rFID from 2.39 to 2.10 on ImageNet-1k while achieving 100% utilization.
-
Product Quantization for Surface Soil Similarity
A pipeline using product quantization and systematic parameter evaluation creates data-driven soil taxonomies with higher specificity than human-derived classifications.
-
Large-Scale Data Parallelization of Product Quantization and Inverted Indexing Using Dask
Dask parallelization of product quantization and inverted indexing allows large-scale approximate nearest neighbor search while preserving accuracy and reducing computation to medium-scale levels.
- Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations