[WebGPU] Fix MHA to ignore past key/value when no present outputs requested

[vraspar]

Description

When MultiHeadAttention has only 1 output (no present_key/present_value outputs), past key/value inputs should be completely ignored, matching CPU EP semantics. The WebGPU EP was passing pastKey/pastValue TensorViews to shader creation functions even when outputCount <= 1, which affected shader cache keys and allowed past data to leak into the attention computation.

This caused the test "MultiHeadAttention Basic, one head and head-size=4 with pastKey and pastValue" to fail with output [17,18,19,20] (pastValue data) instead of expected [9,10,11,12] (V data). The failing output matches exactly what happens when past IS used: Q·pastKey=75 dominates Q·K=35, so softmax gives ~100% weight to pastValue.

Fix

In applyAttention(), introduce effectivePastKey/effectivePastValue that are set to undefined when outputCount <= 1. All downstream usage (shader creation, input arrays) uses these effective values instead of the raw parameters. This ensures:

• Shader cache keys correctly reflect the "no past" configuration
• Past tensors are never passed to any shader creation function
• Behavior matches CPU EP (which ignores past when present outputs are null)
• GQA is unaffected (always has outputCount >= 3)
• Vanilla Attention is unaffected (always passes undefined for past)

[Copilot]

Pull request overview

This PR fixes WebGPU EP MultiHeadAttention behavior to match CPU EP semantics by ensuring past key/value inputs are ignored when the node requests only the primary output (i.e., no present_key / present_value outputs). This prevents past KV from influencing attention results and avoids incorrect shader cache key specialization when present outputs aren’t requested.

Changes:

• Introduce effectivePastKey / effectivePastValue that are set to undefined when outputCount <= 1.
• Route all downstream usage (program creation + input binding) through the effective past values so past tensors are never passed to shader creation/dispatch in the “no present outputs” case.

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.