4 papers
Transformer Key-Value Memories Are Nearly as Interpretable as Sparse Autoencoders
Mengyu Ye, Jun Suzuki, Tatsuro Inaba +1
Recent interpretability work on large language models (LLMs) has been increasingly dominated by a feature-discovery approach with the help of proxy modules. Then, the quality of fe…
How a Bilingual LM Becomes Bilingual: Tracing Internal Representations with Sparse Autoencoders
Tatsuro Inaba, Go Kamoda, Kentaro Inui +5
This study explores how bilingual language models develop complex internal representations. We employ sparse autoencoders to analyze internal representations of bilingual language…
TopK Language Models
Ryosuke Takahashi, Tatsuro Inaba, Kentaro Inui +1
Sparse autoencoders (SAEs) have become an important tool for analyzing and interpreting the activation space of transformer-based language models (LMs). However, SAEs suffer severa…
Weight-based Analysis of Detokenization in Language Models: Understanding the First Stage of Inference Without Inference
Go Kamoda, Benjamin Heinzerling, Tatsuro Inaba +3
According to the stages-of-inference hypothesis, early layers of language models map their subword-tokenized input, which does not necessarily correspond to a linguistically meanin…