Xiang Ren

Examining machine and human reasoning in language processing tasks.

A study reveals gaps in LLMs' understanding of logic rules compared to humans.

A new method reduces forgetting in language models during updates.

This method enhances transparency in determining document authorship.

A new method to assess long-context language models’ learning abilities through Task Haystack.

Enhancing LLMs with external memory aids multi-step reasoning tasks.

A tool to analyze chat logs quickly and effectively for researchers.

A hybrid system combines human insights with machine forecasts for improved decision-making.

Examining the impact of cultural bias in language models and the need for diverse representation.