Entanglement Fine-Tuning in Composite Quantum Systems

Peter W. Evans

Despite attempts to apply causal modeling techniques to quantum systems, Wood and Spekkens argue that any causal model purporting to explain quantum correlations must be fine tuned; it must violate the assumption of faithfulness. This paper is an attempt to undermine the reasonableness of the assumption of faithfulness in the quantum context. Employing a symmetry relation between an entangled quantum system and a “sideways” quantum system consisting of a single photon passing sequentially through two polarizers, I argue that Wood and Spekkens’s analysis applies equally to this sideways system also. As a result, we must either reject a causal explanation in this single photon system, or the sideways system must be fine tuned. If the latter, a violation of faithfulness in the ordinary entangled system may be more tolerable than first thought. Thus, extending the classical “no fine-tuning” principle of parsimony to the quantum realm may be too hasty.

Artur Sowa

Stefan Schumacher, Jens Förstner, A. Zrenner +4 more

We study the quantum properties and statistics of photons emitted by a quantum-dot biexciton inside a cavity. In the biexciton-exciton cascade, fine-structure splitting between exciton levels degrades polarization-entanglement for the emitted pair of photons. However, here we show that the polarization-entanglement can be preserved in such a system through simultaneous emission of two degenerate photons into cavity modes tuned to half the biexciton energy. Based on detailed theoretical calculations for realistic quantum-dot and cavity parameters, we quantify the degree of achievable entanglement.

G. C. Ghirardi, Luca Marinatto, Tullio Weber

Katherine Ip, Casey R. Myers, Udaya Parampalli +2 more

Large language models (LLMs) exhibit abilities beyond natural language modelling and text generation. Recent advances in their reasoning capabilities have spurred interest in applying LLMs to complex scientific tasks requiring deep domain expertise and sophisticated reasoning. Quantum computing, as a highly specialised field with significant knowledge barriers and hardware constraints, could greatly benefit from such advancements. However, a key open question that first must be answered is: How can we develop fine-tuning pipelines that instil genuine quantum reasoning in LLMs, rather than task-specific pattern matching? We study this question through quantum circuit simulation as a training objective, where the model must predict the measurement probability distribution resulting from a sequence of quantum gate operations. We propose and compare two fine-tuning pipelines: (1) Supervised Fine-Tuning (SFT) on explicit gate-by-gate state-vector simulation traces, and (2) a two-stage SFT+Group Relative Policy Optimisation (GRPO) approach that sequentially applies SFT followed by GRPO with verifiable rewards. Our findings show that SFT achieves near-perfect in-distribution and gate-count extrapolation accuracy, significantly outperforming both the base model and the GPT-OSS-120B baseline. SFT+GRPO trades some in-distribution precision for better generalisation to larger qubit systems that SFT alone cannot handle. Both pipelines significantly outperform the baselines, demonstrating that targeted fine-tuning on explicit reasoning traces is an effective strategy for advancing quantum reasoning in LLMs.

V. M. Akulin