Gemma-2

Gemma2 is a dataset used to evaluate the performance of models in the context of sequential knowledge editing and their ability to maintain integrity after multiple edits.

Papers using Gemma-2 (5)

SCAN: Sparse Circuit Anchor Interpretable Neuron for Lifelong Knowledge Editing2026Decomposing MLP Activations into Interpretable Features via Semi-Nonnegative Matrix Factorization2025Precise In-Parameter Concept Erasure in Large Language Models2025Steering Large Language Model Activations in Sparse Spaces2025Franken-Adapter: Cross-Lingual Adaptation of LLMs by Embedding Surgery2025