SOURCE-LINKED INTELLIGENCE

Weakening Neurons: An Input-Output Functionality in Transformers with Outsize Influence

arXiv · AI, language, vision and robotics · article · Sep 16, 2026 · UTC

We analyze the learned input-output behavior of GLU-based neurons in large language models (LLMs). We propose a simple analysis method: For each neuron, we compute the cosine similarities between its input (reading) and output (writing) weight vectors. In this scheme, a strong negative cosine similarity indicates the neuron weakens the direction it detects in the residual stream, so we call this a weakening neuron. This allows us to gain a number of novel insights. First, we show that nine different LLMs have similar patterns: weakening neurons appear mostly in late layers whereas their counte

Read original source ↗ Open in workspace

recordType

paper

region

Global