bytez
Search
Feed
Models
Agent
Devs
API Dashboard
docs
GitHub
ProxySPEX: Inference-Efficient Interpretability via Sparse Feature Interactions in LLMs | Read Paper on Bytez
ProxySPEX: Inference-Efficient Interpretability via Sparse Feature Interactions in LLMs
2 weeks ago
·
arXiv