Exploring Model Stability of Deep Neural Networks for Reliable RRAM-Based In-Memory Acceleration

Mathematics

• Neural Networks 67%
• Pruning 30%
• Model 23%
• Hardware 19%
• Compression 16%
• Cross-layer 15%
• Quantization 15%
• Wafer 12%
• Experimental Evaluation 11%
• Chip 10%
• Stable Models 9%
• Computing 9%
• Neural Network Model 9%
• Performance 9%
• Architecture 8%
• Accelerate 8%
• Training 8%
• Tolerance 8%
• Sparsity 8%
• High Accuracy 8%
• Model Selection 7%
• Benchmark 7%
• Maximise 7%
• Necessary 5%
• Metric 5%

Engineering & Materials Science

• RRAM 100%
• Deep neural networks 69%
• Data storage equipment 43%
• Computer hardware 8%