This paper presents a unified framework for model editing that unifies two popular methods, ROME and MEMIT, under a single conceptual umbrella. The key insight is that both methods optimize the same objective function, known as the preservation-memorization objective. ROME uses an equality constraint to optimize this objective for single edits, while MEMIT uses a more flexible least-square constraint that allows for batched edits. The authors introduce EMMET, a new algorithm that enables batched editing with equality constraints, achieving performance similar to MEMIT across multiple dimensions. EMMET allows for batch sizes up to 10,000 and enables the true unification of ROME and MEMIT, showing that both algorithms are equivalent in terms of their optimization objective, their abilities (singular and batched editing), their model editing performance, and their limitations. The paper also discusses the impact of edit-distribution algorithms on model performance, showing that while MEMIT performs slightly better for Llama-2-7b, EMMET performs slightly better when using the MEMIT edit-distribution algorithm. The results indicate that both EMMET and MEMIT degrade model performance similarly, suggesting that the preservation-memorization objective may have reached its limits. The paper also highlights the importance of understanding edit distribution and its implications for model editing.This paper presents a unified framework for model editing that unifies two popular methods, ROME and MEMIT, under a single conceptual umbrella. The key insight is that both methods optimize the same objective function, known as the preservation-memorization objective. ROME uses an equality constraint to optimize this objective for single edits, while MEMIT uses a more flexible least-square constraint that allows for batched edits. The authors introduce EMMET, a new algorithm that enables batched editing with equality constraints, achieving performance similar to MEMIT across multiple dimensions. EMMET allows for batch sizes up to 10,000 and enables the true unification of ROME and MEMIT, showing that both algorithms are equivalent in terms of their optimization objective, their abilities (singular and batched editing), their model editing performance, and their limitations. The paper also discusses the impact of edit-distribution algorithms on model performance, showing that while MEMIT performs slightly better for Llama-2-7b, EMMET performs slightly better when using the MEMIT edit-distribution algorithm. The results indicate that both EMMET and MEMIT degrade model performance similarly, suggesting that the preservation-memorization objective may have reached its limits. The paper also highlights the importance of understanding edit distribution and its implications for model editing.