An LLM multi-agent framework for paper-to-code generation succeeded on all eight benchmarked ML papers and reached 49.2% on PaperBench Code-Dev, ahead of prior systems.
Entropy Aware Message Passing in Graph Neural Networks
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
abstract
Deep Graph Neural Networks struggle with oversmoothing. This paper introduces a novel, physics-inspired GNN model designed to mitigate this issue. Our approach integrates with existing GNN architectures, introducing an entropy-aware message passing term. This term performs gradient ascent on the entropy during node aggregation, thereby preserving a certain degree of entropy in the embeddings. We conduct a comparative analysis of our model against state-of-the-art GNNs across various common datasets.
fields
cs.SE 1years
2025 1verdicts
CONDITIONAL 1representative citing papers
citing papers explorer
-
AutoP2C: An LLM-Based Agent Framework for Code Repository Generation from Multimodal Content in Academic Papers
An LLM multi-agent framework for paper-to-code generation succeeded on all eight benchmarked ML papers and reached 49.2% on PaperBench Code-Dev, ahead of prior systems.