Summing Up the Facts: Additive Mechanisms Behind Factual Recall in LLMs
This addresses the problem of understanding knowledge retrieval mechanisms in LLMs for researchers, but it is incremental as it builds on existing mechanistic analysis methods.
The paper investigates how transformer-based large language models store and retrieve knowledge, specifically in factual recall tasks, finding that multiple independent mechanisms additively combine to produce correct answers through constructive interference.
How do transformer-based large language models (LLMs) store and retrieve knowledge? We focus on the most basic form of this task -- factual recall, where the model is tasked with explicitly surfacing stored facts in prompts of form `Fact: The Colosseum is in the country of'. We find that the mechanistic story behind factual recall is more complex than previously thought. It comprises several distinct, independent, and qualitatively different mechanisms that additively combine, constructively interfering on the correct attribute. We term this generic phenomena the additive motif: models compute through summing up multiple independent contributions. Each mechanism's contribution may be insufficient alone, but summing results in constructive interfere on the correct answer. In addition, we extend the method of direct logit attribution to attribute an attention head's output to individual source tokens. We use this technique to unpack what we call `mixed heads' -- which are themselves a pair of two separate additive updates from different source tokens.