Clarification on Differences Between Influence Approximation in Section 3.2 and "Estimating Training Data Influence by Tracing Gradient Descent"

[wanweilin]

Hello,

I am currently studying the method described in Section 3.2, "Influence Approximation", of your paper and I am intrigued by its approach and potential applications. However, I came across a similar concept in the paper titled "Estimating Training Data Influence by Tracing Gradient Descent" and I'm trying to understand the distinctions between the two methodologies.

Could you please clarify the differences between the influence calculation method proposed in your section 3.2 and the one introduced in "Estimating Training Data Influence by Tracing Gradient Descent"?

Thank you for your time and consideration.

Best regards.