Transformer Massive Activations and Attention Sinks Share Same Root

alfcnz · x · 2026-07-09

This thread discusses two common phenomena in Transformers: massive activations in a few residual channels, and abnormal attention concentration on a specific "sink" token. The author claims that an ICML 2026 paper proves both stem from the same mechanism: early FFNs inject massive spikes that subsequent layers later cancel out. RMSNorm then forces the spiked token into an approximately fixed direction, making it a sink token that all queries easily attend to. The thread also explains the practical implications: massive activations break quantization because outliers shrink normal values, while attention sinks are useful for streaming inference, as fixing these sink tokens in the KV cache allows the model to run longer.

Related event: ICML 2026 to Explore Massive Activations and Attention Sinks in Transformers(2 posts)→

Original post →

More from Research

Research channel →