# `Edifice.Blocks.CrossAttention`
[🔗](https://github.com/blasphemetheus/edifice/blob/main/lib/edifice/blocks/cross_attention.ex#L1)

Cross-Attention Layer.

Standard encoder-decoder attention where queries come from one sequence
and keys/values come from another. Used in U-Net conditioning, Perceiver,
CLIP text-image alignment, and sequence-to-sequence models.

## Architecture

```
Query source [batch, seq_q, dim_q]    KV source [batch, seq_kv, dim_kv]
      |                                      |
   Dense Wq                           Dense Wk, Wv
      |                                  |      |
      Q                                  K      V
      |                                  |      |
      +----- Attention(Q, K, V) ---------+------+
                    |
             Dense Wo (output projection)
                    |
Output [batch, seq_q, hidden_size]
```

## Usage

    output = CrossAttention.layer(queries, context,
      hidden_size: 256,
      num_heads: 4,
      name: "cross_attn"
    )

## References
- "Attention Is All You Need" (Vaswani et al., 2017)

# `layer`

```elixir
@spec layer(Axon.t(), Axon.t(), keyword()) :: Axon.t()
```

Build a cross-attention Axon layer.

## Parameters
  - `query_input` - Query sequence Axon node [batch, seq_q, dim_q]
  - `kv_input` - Key-value sequence Axon node [batch, seq_kv, dim_kv]

## Options
  - `:hidden_size` - Hidden dimension for Q, K, V projections (required)
  - `:num_heads` - Number of attention heads (default: 1)
  - `:dropout` - Dropout rate (default: 0.0)
  - `:name` - Layer name prefix (default: "cross_attn")

---

*Consult [api-reference.md](api-reference.md) for complete listing*