AI RESEARCH
Lost in the Middle at Birth: An Exact Theory of Transformer Position Bias
arXiv CS.AI
•
ArXi:2603.10123v1 Announce Type: cross The ``Lost in the Middle'' phenomenon -- a U-shaped performance curve where LLMs retrieve well from the beginning and end of a context but fail in the middle -- is widely attributed to learned Softmax artifacts or the distance-decay of positional encodings like RoPE. This paper makes a single, precise claim: \emph{the U-shape is already present at initialization, before any