Concatenate the heads and project to the residual width

Read the equation

The 128 heads contribute 128 value components each, giving 16,384 components per token. The output linear map then projects that concatenation into the 7,168-dimensional residual stream. The indicated tensor shape describes the weight layout; executing the quantized checkpoint requires the corresponding loading and dequantization path.

Terms & definitions

How it works

Read the definition from its right-hand expression to the quantity on the left. Its surrounding source records the intended input, state, or parameter context. Where the source leaves an operation unspecified, this page preserves that gap.

Assumptions & limits

Source recovery establishes that the expression appears in this record. Its mathematical validity, implementation behavior, and interpretation must be assessed separately. See the linked source and research notes for the checks available for this version.

Original context

scores = q @ k^T * 0.135237
      probs  = softmax(scores, dim=-1) computed in float32, cast back to q.dtype
      out    = probs @ v                            -> 128 per head
      out    = o_proj(out.reshape(T, 16384))        -> 7168  fp8 (7168, 16384)

  WHAT IS CACHED - THE POINT OF MLA
      A conventional 128-head KV cache would store K and V per head:
          128 heads x (128 + 128) x 2 bytes = 65,536 B per token per layer.

out    = o_proj(out.reshape(T, 16384))        -> 7168  fp8 (7168, 16384)

Related builds