2202.05262
arxiv.org · 7,863 words · saved by 1 readers
N/A
Locating and Editing Factual Associations in GPT Kevin Meng∗ David Bau∗ Alex Andonian Yonatan Belinkov† MIT CSAIL Northeastern University MIT CSAIL Technion – IIT Abstract arXiv:2202.05262v5 [cs.CL] 13 Jan 2023 We analyze the storage and recall of factual associations in…
related reading
- But is it really in Rome? An investigation of the ROME model editing technique — AI Alignment Forumalignmentforum.org
- On the Biology of a Large Language Modeltransformer-circuits.pub
- Circuit Tracing: Revealing Computational Graphs in Language Modelstransformer-circuits.pub
- Believe It or Not: How Deeply do LLMs Believe Implanted Facts?alignment.anthropic.com
- Mass-Editing Memory in a Transformerarxiv.org
- Fact Finding: Attempting to Reverse-Engineer Factual Recall on the Neuron Level (Post 1) — AI Alignment Forumalignmentforum.org
- [2510.17941] Believe It or Not: How Deeply do LLMs Believe Implanted Facts?arxiv.org
- [2005.11401] Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasksarxiv.org
- Can a Language Model Learn Facts Continually in Its Weights?labs.baseten.co
- 🧠 MLPs are Hebbian Memories: A Simple Recipe for Fact-Storing Transformershazyresearch.stanford.edu
- Thinking to recall: How reasoning unlocks parametric knowledge in LLMsresearch.google
- Understanding Memorization via Loss Curvaturegoodfire.ai