RISE applies CountSketch to dual lexical and semantic channels derived from output-layer gradient outer products, cutting data attribution storage by up to 112x and enabling retrospective and prospective influence analysis on LLMs up to 32B parameters.
Feature hashing for large scale multitask learning
3 Pith papers cite this work. Polarity classification is still indexing.
3
Pith papers citing it
abstract
Empirical evidence suggests that hashing is an effective strategy for dimensionality reduction and practical nonparametric estimation. In this paper we provide exponential tail bounds for feature hashing and show that the interaction between random subspaces is negligible with high probability. We demonstrate the feasibility of this approach with experimental results for a new use case -- multitask learning with hundreds of thousands of tasks.
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
TrapNet applies PCA-based FloatHash vectors and graph community detection to enable unsupervised malware fingerprinting and family attribution from static analysis.
A neural semantic matcher for product search uses a custom loss on behavior data, n-gram pooling, and hashing to beat prior methods by 4.7% Recall@100 and 14.5% MAP.
citing papers explorer
-
Sketching the Readout of Large Language Models for Scalable Data Attribution and Valuation
RISE applies CountSketch to dual lexical and semantic channels derived from output-layer gradient outer products, cutting data attribution storage by up to 112x and enabling retrospective and prospective influence analysis on LLMs up to 32B parameters.
-
Applying Graph Analysis for Unsupervised Fast Malware Fingerprinting
TrapNet applies PCA-based FloatHash vectors and graph community detection to enable unsupervised malware fingerprinting and family attribution from static analysis.
-
Semantic Product Search
A neural semantic matcher for product search uses a custom loss on behavior data, n-gram pooling, and hashing to beat prior methods by 4.7% Recall@100 and 14.5% MAP.