Jonas Loos
JonasLoos
AI & ML interests
Interpretability
Recent Activity
authored a paper about 3 hours ago
KV-Cache Patching: Tracing How and Where Fine-Tuned Behaviors Arise updated a collection about 4 hours ago
KV-Cache Patching published a model about 4 hours ago
JonasLoos/Qwen3.5-4B-kv-patching