Token-wise Influential Training Data Retrieval for Large Language Models

​In this session, we welcome Huawei Lin from the Rochester Institute of Technology, who co-authored the paper "Token-wise Influential Training Data Retrieval for Large Language Models".

​The paper can be found here.

​

​The paper proposes a scalable framework adapting to LLMs to estimate the influence of
each training data on the LLM output.