Maximum normalized projection-body volume

A projection body encodes the volumes of a convex body’s projections onto hyperplanes. Brannen asked whether a simplex maximizes the volume of this new body after normalizing by the original body’s volume.

Why it matters and what counts as progress

Why it matters: Normalization makes the comparison invariant under changes of scale and affine coordinates. Explicit examples test which shapes can be extremal.

What counts as progress: Prove a dimension-specific upper bound with its equality cases, or construct a body whose normalized projection volume exceeds the simplex value. Use the projection body itself in the comparison.

How the problem developed

  1. 1996 · Simplex maximum question [1]

    N. S. Brannen Proposed that a simplex maximizes normalized projection-body volume.

  2. 21 Aug 2026 · Earlier counterexamples [2]

    Yibin Feng, Shengnan Hu, Weiru Liu and Lei Xu Constructed counterexamples in every dimension at least nine.

  3. 6 Oct 2026 · Reported product counterexample [3]

    OpenAI’s internal model The Cartesian product of two ten-dimensional simplices has larger normalized projection-body volume than a twenty-dimensional simplex, with the ratio given exactly in the paper.

AI contribution

Sources

  1. Volumes of projection bodies
  2. On the Reverse Projection Inequality
  3. A product counterexample to the simplex maximum for projection-body volume
  4. OpenAI mathematics release
  5. OpenAI’s Lean scope for this manuscript
Publication history