Skip to content

Convert Claude Code or Codex into the most intelligent Deep Research Agent. Collect, search, and synthesize web research into a persistent, searchable wiki that builds on itself. Hosted API + MCP: hyperresearch.ai

docs: benchmark chart — truncate y-axis at 30 for readability (#5)

Jordan Gibbs725edb5bdfe0
diff --git a/assets/_generate_benchmark.py b/assets/_generate_benchmark.py
index f4049e2..e562ced 100644
--- a/assets/_generate_benchmark.py
+++ b/assets/_generate_benchmark.py
@@ -114,15 +114,17 @@ for i, tick in enumerate(ax.get_xticklabels()):
tick.set_fontweight("bold")
tick.set_fontsize(13)
-# Y-axis
+# Y-axis — truncated at 30 so differentiation in the 45-60 range reads
+# clearly. Scores in this regime are compressed; 0-100 full scale would
+# flatten everything into identical-looking bars.
ax.set_ylabel(
- "RACE overall score (0–100, higher = better)",
+ "RACE overall score (axis truncated at 30; max = 100)",
color=dim,
fontsize=11,
labelpad=10,
)
ax.tick_params(axis="y", colors=dim, labelsize=10)
-ax.set_ylim(0, max(scores) * 1.18) # headroom for star + labels
+ax.set_ylim(30, max(scores) * 1.12) # headroom for star + labels
# Gridlines
ax.yaxis.grid(True, color=grid, linewidth=0.8, zorder=1)
diff --git a/assets/benchmark.png b/assets/benchmark.png
index 451e9dc..3aaaaf3 100644
Binary files a/assets/benchmark.png and b/assets/benchmark.png differ