Biswajit Das, Sreekanth K. Manikandan and Ayan Banerjee (2025), Physical Review Applied 24, 034039.
Key takeaways
From the reading, NOTE-788.
- A known causal structure. The coupling is symmetric, H₁₂ = H₂₁ = ε/γ, with ε = 3a/(2d) − (a/d)³. Only trap 1 is driven, and the drive λ is recorded.
- The measures recover the order. At θ = 0.56 and ε = 0.49, in experiment and numerics, MI_{λ→x₁}(τ) > MI_{x₁→λ}(τ) and MI_{x₁→x₂}(τ) > MI_{x₂→x₁}(τ). Transfer entropy agrees, with TE_{λ→x₁} ≫ TE_{x₁→λ}.
- But the flow between the particles is two-way at moderate drive. The reverse measures also peak (λ → x₁ ⇄ x₂). Only at θ = 5.6, in simulation, does the flow become one-way.
- Coupling redistributes information. A larger ε lowers MI_{λx₁} and raises MI_{x₁x₂} and MI_{λx₂}. The last two are too small to measure in the experiment’s regime and are given from numerics.
- The irreversibility “explanation” is a parallel of trends. TE asymmetry and partial entropy production vary alike with θ and ε. No derivation links them.
- With no drive, the coupled particles are independent. P(x₁, x₂) = P(x₁)P(x₂), and the delayed measures are symmetric.
Standing in the record
Filed on 2026-10-12 for QUESTION-063 and read on 2026-10-14 in NOTE-788. The registration said its measures “tell an external drive from coupling”. They do not. The drive is applied to one particle and recorded, so the measures locate where it enters. No unrecorded drive shared by both particles, the case QUESTION-063 is about, is studied. What the paper does establish bears on the question as a limit. The direction the measures report follows the drive and its strength, while the coupling stays symmetric. That is filed as THEORY-275, which answers QUESTION-063 in part.
It also qualifies CLAIM-172 and CLAIM-167. A mutually coupled pair reads as a “hierarchical interaction” once one member is driven hard, so observational flow cannot tell peers from hierarchy without an intervention on a member.