Upload REPORT.md with huggingface_hub

REPORT.md

@@ -365,7 +365,7 @@ at one R.
   not a capacity, failure
 - prediction (a) **not supported**
 - prediction (b) **not supported**
-- **100k-step probe (1 seed, 4x budget)**: id acc 0.235, k=16 0.035, k=64 0.015 -> **not merely step-starved**: 4x the steps did not move the plateau, so something about optimising the variable-depth readout is fundamentally harder at this scale (lr, zero-init dynamics, or gradient conflict across depths)
+- **100k-step probe (1 seed, 4x budget)**: id acc 0.235, k=16 0.035, k=64 0.015 -> **a slow learner that plateaus**: it climbed 0.10 -> 0.235 by 60k steps, then stayed flat for the last 40k (slope ~0), so the aligned objective is roughly 4-10x slower to optimise than fixed-R training and did not break out of its second plateau by 100k. Why: with supervision only at the end of the chain, the model faces a credit-assignment problem across depths - it must learn to do one hop per unroll AND to stop exactly at depth k, with no per-step signal telling it either. Even the easiest case (k=1: one unroll) stayed unsolved at 25k steps.
 
 ## 5. Conclusions