Whether DSP_2 contains all PPT maps

Determine whether every PPT map on M_d, for d≥3, belongs to the cone DSP_2 of sums of a 2-superpositive map and the transpose of a 2-superpositive map; equivalently, establish whether every 2-positive and 2-copositive map is decomposable.

Background

The paper introduces DSP_k as the cone formed by adding the k-superpositive cone to its transpose image. It proves strong entanglement-breaking composition results for PPT maps interacting with DSP_2, but explicitly states that it is unknown whether DSP_2 contains every PPT map when d≥3. An affirmative answer would imply the relevant PPT-cubed conjecture.

References

Furthermore, to the best of our knowledge, it is unknown whether the cone $DSP_2$ contains all PPT maps when $d\geq 3$.

Every PPT channel has finite entanglement-breaking index  (2608.13551 - Park, 13 Aug 2026) in Section 1, Introduction, subsection 'Absolute bounds on the EB index via low-entanglement decompositions'

Indeed, this assertion is equivalent to the dual statement \begin{equation} \label{eq:2biPTDec} \Phi \text{ is $2$-positive and $2$-copositive} \stackrel{?}{\implies} \Phi\in DEC, \end{equation} and even a stronger statement posed in , whether every $2$-entanglement-breaking map is decomposable, remains open.

Every PPT channel has finite entanglement-breaking index  (2608.13551 - Park, 13 Aug 2026) in Section 1, Introduction, subsection 'Absolute bounds on the EB index via low-entanglement decompositions'

We leave open the following question.

\begin{question} What is the smallest integer $k\leq d$ such that $$DSP_k\supseteq PPT?$$ \end{question}

Every PPT channel has finite entanglement-breaking index  (2608.13551 - Park, 13 Aug 2026) in Section 5, 'Entanglement-breaking-composability via low degree of entanglement'