MIP*=RE
Zhengfeng Ji, Anand Natarajan, Thomas Vidick, John Wright, Henry Yuen
quant-ph
Jan 13, 2020 · v4
cs.CC math.OA
TL;DR
The authors used general LLM methods to formalize the paper's main statement (Corollary 12.9) in Lean.
Abstract
We show that the class MIP* of languages that can be decided by a classical verifier interacting with multiple all-powerful quantum provers sharing entanglement is equal to the class RE of recursively enumerable languages. Our proof builds upon the quantum low-degree test of (Natarajan and Vidick, FOCS 2018) and the classical low-individual degree test of (Ji, et al., 2020) by integrating recent developments from (Natarajan and Wright, FOCS 2019) and combining them with the recursive compression framework of (Fitzsimons et al., STOC 2019). An immediate byproduct of our result is that there is an efficient reduction from the Halting Problem to the problem of deciding whether a two-player nonlocal game has entangled value $1$ or at most $1/2$. Using a known connection, undecidability of the entangled value implies a negative answer to Tsirelson's problem: we show, by providing an explicit example, that the closure $C_{qa}$ of the set of quantum tensor product correlations is strictly included in the set $C_{qc}$ of quantum commuting correlations. Following work of (Fritz, Rev. Math. Phys. 2012) and (Junge et al., J. Math. Phys. 2011) our results provide a refutation of Connes' embedding conjecture from the theory of von Neumann algebras.
Problem
The paper asks whether MIP*, the class of languages decidable by a classical verifier interacting with multiple entangled quantum provers, equals RE, the class of recursively enumerable languages. The question bears on Tsirelson's problem and Connes' embedding conjecture.
Approach
The proof combines the quantum low-degree test, the classical low-individual-degree test, and introspection techniques. These are integrated into a recursive compression framework for normal form verifiers that use conditionally linear question distributions. The main statement (Corollary 12.9) was formalized in Lean using general LLM methods.
Results
The authors prove MIP* = RE, which gives an efficient reduction from the Halting Problem to approximating the entangled value of nonlocal games. Consequences are that C_qa is strictly contained in C_qc, a negative answer to Tsirelson's problem, and a refutation of Connes' embedding conjecture.