KARMA updates advertising-agent knowledge without changing model weights
Authors report average hold-out score rising from 30.19 to 39.15 across 18 domain–model settings.
Stories per day · 30 days
4 stories, newest first · one card per story
Stories tagged #knowledge-editing
EngramEdit: Decoupled Knowledge Updates in LLMs through Conditional Memory
EngramEdit updates facts in DeepSeek Engram conditional memory while keeping the Transformer backbone fixed.
Improving Atomic-Fact Recall via Focused Views in Unstructured Knowledge Editing
FOVEATED shifts RoPE context during editing so models recall individual facts from unstructured passages.
The Tokens Remember: When Tokenization Bypasses Knowledge Editing and Unlearning
Toketive shows alternative tokenizations can bypass LLM knowledge edits and recover suppressed answers.