This is a brief message of appreciation to @bernardodionisi and contributors for developing this library.
We use (and cite) this software in our latest paper on memorisation in large language models (https://arxiv.org/pdf/2406.04327). In the paper, we use the method proposed by Callaway and Sant'Anna (2021) to estimate memorisation in language models, which implies running their method on a big panel. We tried other implementations (mainly in R); however, differences was the only one able to scale out-of-the-box to our panel size (even using the bootstrapped standard errors option).
Overall, this library has made my work so much easier, and I hope it will continue improving in the future!
Hi there,
This is a brief message of appreciation to @bernardodionisi and contributors for developing this library.
We use (and cite) this software in our latest paper on memorisation in large language models (https://arxiv.org/pdf/2406.04327). In the paper, we use the method proposed by Callaway and Sant'Anna (2021) to estimate memorisation in language models, which implies running their method on a big panel. We tried other implementations (mainly in R); however,
differences
was the only one able to scale out-of-the-box to our panel size (even using the bootstrapped standard errors option).Overall, this library has made my work so much easier, and I hope it will continue improving in the future!
Best, Pietro