Wave 7: DeepSeek/RNN (paper96-110)
Integration of DeepSeek R2 and RNN architectures with spiking neural networks. Coordinated by the deepseek-snn-agent.
Projects
| Project | Directory | Purpose |
|---|---|---|
| DeepSpike-R2 | Papers/paper96/ | DeepSeek R2 with SNN components |
| SpikeRNN-Hybrid | Papers/paper97/ | Hybrid RNN-SNN models |
| NeuroSeek-Reason | Papers/paper98/ | Reasoning via SNN-enhanced DeepSeek |
| SpikeOCR-Latent | Papers/paper99/ | Latent OCR with spike encoding |
| DeepSpike-Context | Papers/paper100/ | Long-context processing |
| RNN-SNN-Bridge | Papers/paper101/ | Bridging RNN and SNN paradigms |
| NeuroSeek-Vision | Papers/paper102/ | Visual understanding |
| SpikeLatent-Compress | Papers/paper103/ | Latent space compression |
| DeepSpike-Long | Papers/paper104/ | Long-sequence processing |
| RNN-TB-Scale | Papers/paper105/ | Scaling RNN-SNN to TB |
| NeuroSeek-Multi | Papers/paper106/ | Multi-modal integration |
| SpikeRNN-Memory | Papers/paper107/ | Memory mechanisms |
| DeepSpike-Efficient | Papers/paper108/ | Efficiency optimization |
| RNN-SNN-Train | Papers/paper109/ | Training methodology |
| NeuroSeek-Deploy | Papers/paper110/ | Deployment strategies |
All 15 projects are in "Reviewed" status with CLAUDE.md agent context files.