PaperScope
LIVE · 2026-09-09 05:40 UTC

RoPE attention is an exact forward-pass gradient step with softmax intact

Julie Huang, Maggie Chlon, Leon Chlon

Latestcs.CLcs.LGcs.AIcs.CV
arXiv ID
2609.06685 v1
Category
Submitted
2026-09-06

Abstract

We derive an exact gradient-step representation of the RoPE-softmax forward pass. For every deterministic RoPE-softmax attention head with arbitrary affine projection weights, we construct a query-dependent effective matrix $ΔM_i$ satisfying $y_i = μ_i + u_i^\top ΔM_i$, where $μ_i$ is the uniform mean of the attended values and $u_i$ is the augmented query input. The construction applies the classical exponential divided difference $ρ= φ_1$ to retain the softmax exactly. Its positive coefficients give a unit gradient-step representation on a query-conditioned quadratic objective. The same function connects the RoPE generator to exact positional finite differences. We derive a tokenwise formula for the error of reusing one query's matrix and prove that a nonconstant finite-cache head cannot admit a globally exact affine query readout. Reconstruction checks and frozen-reuse calibration on one pretrained Qwen2.5-0.5B layer verify the representation and quantify the correction required when one query's matrix is reused.

arXiv abs page · PDF · same-day batch