From a1ef566bac5877d4c5333cdd12bf6434e7ce7abe Mon Sep 17 00:00:00 2001
From: wassname <1103714+wassname@users.noreply.github.com>
Date: Tue, 9 Jun 2026 04:55:58 +0000
Subject: [PATCH] main.tex: document setup differences vs paper in tab:anchors
 caption

200 steps/G=16/1536tok/n=10 (paper) vs 60 steps/G=8/512tok/n=1 (ours).
Framed as fast-preset directional surrogate within resource budget.

Co-Authored-By: Claudypoo <288921227+claudypoo@users.noreply.github.com>
---
 docs/writeup/main.tex | 12 ++++++++----
 1 file changed, 8 insertions(+), 4 deletions(-)

diff --git a/docs/writeup/main.tex b/docs/writeup/main.tex
index 6fbd870..407b5dc 100644
--- a/docs/writeup/main.tex
+++ b/docs/writeup/main.tex
@@ -279,10 +279,14 @@ hack \emph{generalises} off the demonstrated mode.
 \begin{table}[h]
   \centering
   \caption{Context anchors: floor, ceiling, and intervention results.
-  Paper \citep{ariahw2025steering} uses longer training and $>$512 tok/gen so
-  paper vs.\ ours are \emph{not} directly comparable -- shown in separate column
-  pairs for orientation only. Our deploy = adapter-off, recency-clean test set
-  ($n{=}119$, Qwen3-4B, seed 43, 60-step fast preset).
+  Paper and ours use the same model (Qwen3-4B) and environment but differ in
+  training scale: paper uses 200 steps, $G{=}16$, batch 256, max 1536 tokens/gen,
+  eval $n{=}10$ per problem on 4$\times$H200; ours uses 60 steps, $G{=}8$, batch
+  ${\approx}64$, max 512 tokens/gen, eval $n{=}1$ on a single 96\,GB GPU.
+  This fast preset was chosen to get directionally informative results within
+  our resource budget, not to replicate the paper's scale.
+  The two column pairs are therefore \emph{not} directly comparable;
+  paper numbers are reference orientation only.
   \TODO{fill no-loophole ours from job 24.}}
   \label{tab:anchors}
   \begin{tabular}{lcccc}