Introduces complement-aware submodular functions (CSI) that preserve structure between subset and complement for improved robust data selection.
Submodularity in machine learning and artificial intelligence.ArXiv, abs/2202.00132
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
roles
background 1polarities
background 1representative citing papers
SMA uses a submodular mutual information objective on data sets to deliver competitive zero-shot classification and retrieval performance on CLIP benchmarks with only tens of thousands of samples, orders of magnitude fewer than standard approaches.
A new ordered local search algorithm achieves k/2 + o(k) approximation for monotone submodular maximization over k matroids and (ln 4 k)/3 + o(k) for weighted k-set packing.
Facility location — a classic submodular coverage objective — predicts a training subset's held-out accuracy far better than the Vendi score, which becomes misleading at high values.
An accelerated relax-and-round algorithm for concave coverage problems achieves Õ(mn ε^{-1}) runtime and a 0.827-approximation ratio for the logarithmic reward function.
Bootstrapping math questions via rewriting creates MetaMathQA; fine-tuning LLaMA-2 on it yields 66.4% on GSM8K for 7B and 82.3% for 70B, beating prior same-size models by large margins.
Survey of submodular optimization theory, algorithms, and applications in systems and control, covering properties, constraints, approximations, and control-specific objectives.
citing papers explorer
-
Complement Submodular Information Measures for Balanced and Robust Data Selection
Introduces complement-aware submodular functions (CSI) that preserve structure between subset and complement for improved robust data selection.
-
SMA: Submodular Modality Aligner For Data Efficient Multimodal Learning
SMA uses a submodular mutual information objective on data sets to deliver competitive zero-shot classification and retrieval performance on CLIP benchmarks with only tens of thousands of samples, orders of magnitude fewer than standard approaches.
-
Submodular Maximization over Many Matroids via Ordered Local Search
A new ordered local search algorithm achieves k/2 + o(k) approximation for monotone submodular maximization over k matroids and (ln 4 k)/3 + o(k) for weighted k-set packing.
-
How Much Is a Dataset Worth? Scaling Laws, the Vendi Score, and Matrix Spectral Functions
Facility location — a classic submodular coverage objective — predicts a training subset's held-out accuracy far better than the Vendi score, which becomes misleading at high values.
-
Accelerated Relax-and-Round for Concave Coverage Problems
An accelerated relax-and-round algorithm for concave coverage problems achieves Õ(mn ε^{-1}) runtime and a 0.827-approximation ratio for the logarithmic reward function.
-
MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Models
Bootstrapping math questions via rewriting creates MetaMathQA; fine-tuning LLaMA-2 on it yields 66.4% on GSM8K for 7B and 82.3% for 70B, beating prior same-size models by large margins.
-
Submodular Optimization with Applications to Decision and Control
Survey of submodular optimization theory, algorithms, and applications in systems and control, covering properties, constraints, approximations, and control-specific objectives.
- Recall Is Not Enough: A Reader-Context Diagnostic for Budget-Constrained Retrieval-Augmented Generation