linear programming
PulseAugur coverage of linear programming — every cluster mentioning linear programming across labs, papers, and developer communities, ranked by signal.
2 day(s) with sentiment data
-
New LiFT Framework Uses Linear Programming to Control Transformer Overfitting
Researchers have introduced LiFT, a novel framework for fine-tuning transformer models that utilizes linear programming to control overfitting. This method formulates fine-tuning as a bilevel optimization problem, joint…
-
Audio deepfake model explanations found to be fragile
Researchers have demonstrated that explanations for audio deepfake detection models can be manipulated. By introducing imperceptible perturbations, an adversary can alter the model's attribution heatmaps without changin…
-
Lyapunov Framework Enhances Learning in Weakly-Coupled MDPs
Researchers have developed a novel Lyapunov-based framework to analyze the sample complexity of learning in weakly-coupled Markov decision processes (WCMDPs) and Restless Bandits (RBs). This approach offers a more effic…
-
New Benchmark Suite Evaluates AI Self-Correction in Operations Research
Researchers have developed ORLoopBench, a new benchmark suite designed to evaluate and improve the self-correction and behavioral rationality of AI models in Operations Research (OR). The suite includes OR-Debug-Bench w…
-
New Linear Program Method Enhances Process Conformance Checking Speed
Researchers have developed a new method for process conformance checking by reformulating it as a totally unimodular linear program (LP). This LP approach offers significant speedups for longer process traces with devia…
-
Modal cuts AI inference cold starts by 40x with new GPU techniques
Modal has developed a new method to significantly reduce inference cold start times for AI models. By employing techniques like LP, FUSE, C/R, and CUDA-checkpointing, they achieved a 40x improvement in inference speed. …
-
Researchers develop new training methods for neural networks to improve MILP tractability
Researchers have developed new training regularizers for neural network surrogate models that directly improve their tractability within mixed-integer linear programs (MILPs). These regularizers penalize factors like bi…
-
SOC-ICNN: From Polyhedral to Conic Geometry for Learning Convex Surrogate Functions
Researchers have introduced SOC-ICNN, a novel neural network architecture that expands the representational capabilities beyond classical ReLU-based Input Convex Neural Networks (ICNNs). By generalizing from Linear Prog…