memory: pin the GX10 baseline's final figures

The probe finished. Median is 79.36 s/it across ten timed steps with a
min/max of 79.30 to 79.45, and peak memory is 75.1 of 121.6 GiB by
PyTorch's own max_memory_allocated -- 46 GiB spare rather than the 35 I
estimated from a live free reading, which was counting fragmentation and
the resident model rather than the allocation high-water mark.

Resolved attention backend recorded as flex_attention, read off the loaded
model rather than trusted from the request, which is the check I got wrong
the first time.

persistent-memory.d/2026-09-01-pfi-gx10-onboarding.md

@@ -329,10 +329,13 @@ needed beyond `python3-dev`. This box can train.
 
 ### The number
 
-    pfi-gx10    79.30 79.37 79.33 79.45 79.38 79.35 79.35 s/it  -> MEDIAN 79.35
-                (0.19% spread across seven timed steps; warmup 85.90 then 79.32, discarded)
+    pfi-gx10    MEDIAN 79.36 s/it   min/max 79.30 / 79.45   (10 timed steps, 0.19% spread;
+                warmup 85.90 then 79.32, discarded)
+                peak memory 75.1 / 121.6 GiB -- 46 GiB SPARE, comfortably not memory-bound
+                attn_resolved: flex_attention (matched to the harness, read off the model)
+                FINAL, from the probe's own provenance block on the box
                 -> 604 steps ~= 13.3 h
-                memory 86 / 121 GB, ~35 GB headroom -- NOT memory-constrained
+                peak 75.1 / 121.6 GiB (PyTorch max_memory_allocated) -- 46 GiB spare
     ana-ml2     10.8-15.8 s/it (run-03c.log, the real run)
                 -> 604 steps ~= 2.2-2.7 h