activity
20182026
most citedNeural Spectrahedra and Semidefinite Lifts: Global Convex Optimization of Polynomial Activation Neural Networks in Fully Polynomial-Time

5 citations · 11 across the 6 of their papers we have counts for

collaborators
Showing cs.LGShow all

7 papers · 1 filter

cs.LG2026

KForge: LLM-Driven Cross-Platform Kernel Generation for AI Accelerators

Taras Sereda, Burak Bartan, Ankita Nayak +3

Production inference increasingly targets a heterogeneous mix of accelerators. Agentic pipelines interleave reasoning, tool calls, and multi-agent coordination, each with distinct…

cs.LG2025

KForge: Program Synthesis for Diverse AI Hardware Accelerators

Taras Sereda, Tom St. John, Burak Bartan +3

GPU kernels are critical for ML performance but difficult to optimize across diverse accelerators. We present KForge, a platform-agnostic framework built on two collaborative LLM-b…

cs.LG2023

Moccasin: Efficient Tensor Rematerialization for Neural Networks

Burak Bartan, Haoming Li, Harris Teague +2

The deployment and training of neural networks on edge computing devices pose many challenges. The low memory nature of edge devices is often one of the biggest limiting factors en…

cs.LG2021

Training Quantized Neural Networks to Global Optimality via Semidefinite Programming

Burak Bartan, Mert Pilanci

Neural networks (NNs) have been extremely successful across many tasks in machine learning. Quantization of NN weights has become an important topic due to its impact on their ener…

cs.LG20215 cited

Neural Spectrahedra and Semidefinite Lifts: Global Convex Optimization of Polynomial Activation Neural Networks in Fully Polynomial-Time

Burak Bartan, Mert Pilanci

The training of two-layer neural networks with nonlinear activation functions is an important non-convex optimization problem with numerous applications and promising performance i…

cs.LG20202 cited

Debiasing Distributed Second Order Optimization with Surrogate Sketching and Scaled Regularization

Michał Dereziński, Burak Bartan, Mert Pilanci +1

In distributed second order optimization, a standard strategy is to average many local estimates, each of which is based on a small sketch or batch of the data. However, the local…