Výsledky vyhledávání - "Mahdavi, Sadegh"

Report

From Graph Diffusion to Graph Classification

Autor: Xian, Jia Jun Cheng, Mahdavi, Sadegh, Liao, Renjie, Schulte, Oliver

Generative models such as diffusion models have achieved remarkable success in state-of-the-art image and text tasks. Recently, score-based diffusion models have extended their success beyond image generation, showing competitive performance with dis

Externí odkaz: http://arxiv.org/abs/2411.17236

Zobrazit plný text záznamu

Report

Leveraging Environment Interaction for Automated PDDL Translation and Planning with Large Language Models

Autor: Mahdavi, Sadegh, Aoki, Raquel, Tang, Keyi, Cao, Yanshuai

Large Language Models (LLMs) have shown remarkable performance in various natural language tasks, but they often struggle with planning problems that require structured reasoning. To address this limitation, the conversion of planning problems into t

Externí odkaz: http://arxiv.org/abs/2407.12979

Zobrazit plný text záznamu

Report

Memorization Capacity of Multi-Head Attention in Transformers

Autor: Mahdavi, Sadegh, Liao, Renjie, Thrampoulidis, Christos

Transformers have become the go-to architecture for language and vision tasks, yet their theoretical properties, especially memorization capacity, remain elusive. This paper investigates the memorization abilities of multi-head attention mechanisms,

Externí odkaz: http://arxiv.org/abs/2306.02010

Zobrazit plný text záznamu

Report

Towards Better Out-of-Distribution Generalization of Neural Algorithmic Reasoning Tasks

Autor: Mahdavi, Sadegh, Swersky, Kevin, Kipf, Thomas, Hashemi, Milad, Thrampoulidis, Christos, Liao, Renjie

In this paper, we study the OOD generalization of neural algorithmic reasoning tasks, where the goal is to learn an algorithm (e.g., sorting, breadth-first search, and depth-first search) from input-output pairs using deep neural networks. First, we

Externí odkaz: http://arxiv.org/abs/2211.00692

Zobrazit plný text záznamu

Vyhledávací nástroje:

Upřesnit hledání