[samuelishida]

A native (IQ) pack decodes through verify windows only, so --dump-logits, --dump-layers, and --dump-final-r must emit rows from those windows rather than from the ordinary token loop.

What this PR does:

• Captures the residual ladder inside the verify-window graph when --dump-layers is set. Rows are the fused-layout residual (layer l's write stays pending inside layer l+1's fused read), which is what the engine actually holds.
• Writes --dump-logits and --dump-final-r rows for accepted window rows only; rejected drafts are conditioned on tokens the ordinary path never feeds.
• Rewrites the logits header at close with the real row count so readers see the size the file actually holds.
• Refuses --dump-halves under --spec cleanly.
• Decouples the expert-tier residency-table staging from the dump-gate so the table is available both to the token graph and to verify windows, while preserving the original gate for --no-capture / --no-token-graph paths.
• Provision-gates ple_parity in CMake because its >100 MB fixtures and per-artifact oracle include were never committed; the missing supply chain is documented in tests/python/README.md.

Also documents the native-pack dump semantics in docs/AMD_HIP.md.

[Niko1221]

Thanks @samuelishida, the dump block is now read-only and keeps the verify order, which was our main concern. Still needed for a new PR: the host_res staging condition must stay off for --no-capture / --no-token-graph (it's currently original || (graph_hits && o.spec > 0), which stages under --spec anyway); drop the session-zeroing copy (that part is taken via #167); and the new tests/python/README.md describes a CMake gate and STRATA_PLE_TEST_* options the PR doesn't add. Put the docs in a general doc, not AMD_HIP.md. We're closing PRs that wait on changes so the list stays manageable; this is not a no. Please open a new PR when it's ready and link this one.