aiqbench analysis
This is a theory-level result in quantum computing: a claimed classical simulation of a restricted DQC1 model is challenged by a counterexample built around separable states. For readers, the key point is that the "one clean qubit" model remains an active area where the boundary between classically simulable and genuinely quantum behavior is still being mapped, and that proofs in this area can contain gaps. The work is a specific, peer-review-dependent claim rather than a practical advance in hardware or algorithms.
One clean qubit can control three qubits and stay separable

A new counterexample challenges claims about simulating quantum computers in separable states, using a pure “clean qubit” to control three gates.
Christopher W. Mansell of Terra Quantum UK Ltd. has demonstrated a counterexample to a previously claimed classical simulation result for a specific model of quantum computation. The work centers on Scott Aaronson’s 2005 “semi-grand challenge” questioning whether quantum computers limited to separable mixed states could be efficiently simulated on classical computers. Mansell’s research shows that a single pure qubit, controlling three gates acting on a three-qubit register, can maintain a fully separable state while producing a final unitary incompatible with the earlier simulation claim. This finding reopens the challenge for the actively researched one clean qubit model. Terra Quantum UK Ltd. researchers demonstrated a circuit using one pure qubit to control three qubits while maintaining a fully separable quantum state across all four qubits, a finding that challenges a previously proposed classical simulation result. The team’s work centers on a specific configuration within the deterministic quantum computation with one qubit (DQC1) model, an area of quantum computing that explores the limits of computation with minimal purity. This…
Excerpt shown under the publisher's RSS terms · 167 words · full text and figures are on the publisher's site
Related coverage
Simulated Control Evals
arXiv · Oct 23, 2026
Nvidia’s scale-in play: Controlling agents is the next infrastructure priority
SiliconANGLE · Sep 29, 2026
Suppression Sticks, Locality Is Fragile: A Closed-Loop Target-and-Control Audit of Task-Vector Negation in VLA Policies
arXiv · Aug 6, 2026
What Should an Information Bottleneck Encoder Learn? Separating Representation Learning from Rate Control
arXiv · Oct 28, 2026