The most important result is a non-detection. The full-frequency power spectrum showed no exotic peak at the target frequency above the noise floor. Instead, the preprint reports upper limits on two spin-spin-velocity-dependent interactions, labeled V6 and V14. In ordinary language, the team was looking for a force whose signal would depend on both the spins involved and relative motion. The experiment found no signal strong enough to rise above its stated noise floor, so it set ceilings on how large the two couplings could be under the tested analysis.
A search built around a rhythm
The document describes two individual laboratory experiments, one targeting V6 and the other V14. The setup used a levitated magnet as a force sensor and rotated magnets as spin sources. Magnetic shields were placed between the sensor and the rotated magnets to suppress electromagnetic interactions.
For the V6 protocol, the turntable produced four force peaks during each complete rotation. The interaction frequency was 4ω/(2π), and the search was run at resonance when that frequency aligned with the sensor’s resonant frequency along its z axis. Resonance here means matching the repeating motion to the sensor’s own preferred frequency, so a candidate effect could be sought as a response at a known, controlled frequency.
To check the frequency response, the analysis compared the spectrum from the exotic-force search with a reference spectrum produced by a calibrated coil. It calculated the displacement power spectral density 100 times during a 100-second measurement period and averaged the results. Power spectral density is simply a way of showing how the measured signal’s power is distributed across frequencies. The comparison provided the reference needed to judge whether a feature at the target frequency stood above the instrument’s noise.
The numbers set ceilings
For V6, the reported upper bound on the coupling was |f6| ≤ 2.12 × 10−13 at an interaction range of λ = 1.6 × 10−2 metres, or 1.6 centimetres. The stated constrained range ran from 1 millimetre to 6 centimetres. The paper reports up to a 12-orders-of-magnitude improvement over the previous result. Here, λ is the interaction-range scale used to describe the distance over which the proposed effect was tested.
For V14, the reported upper bound was |f14| ≤ 2.34 × 10−23 at the same λ value, 1.6 × 10−2 metres. Its constrained range extended from 1 millimetre to longer distances, and the reported improvement was 13 orders of magnitude. Both coupling limits were reported at the 95% confidence level. In news terms, they are ceilings from the analysis, not measured values of a force the apparatus observed.
Put in force units, the measurement-noise upper limit was 2.81 × 10−13 newtons for V6 and 4.65 × 10−13 newtons for V14. These values express how much exotic force could remain hidden at the experiment’s noise-limited sensitivity; they are not a claim that either force was detected.
Noise remains the main constraint
The noise result is central to how the numbers should be read. The paper says measurement noise significantly outweighed the other listed noise sources, while subsequent checks found no noise effects at the resonance frequency. The reported relative systematic errors were 32.20% for V6 and 28.97% for V14. The reviewed material also does not spell out the full procedure used to construct the 95% confidence limits.
The result is consequently specific: it tightens constraints on V6 and V14 over the ranges stated in the report, while leaving no target-frequency signal to interpret. The findings are upper limits for the tested interactions, rather than a confirmed observation of an exotic force.
An additional translation of the result gave an upper bound of 3.17 × 10−31 eV−2 for a fundamental coupling combination at λ = 1.6 × 10−2 metres, with up to a 12-orders-of-magnitude improvement over the previous result. That is another upper bound reported from the search, rather than a separate detection.
The paper’s status
The document carries an arXiv version-1 identifier dated 26 Aug 2026. It states that supporting data are openly available. The acknowledgments list support from the National Natural Science Foundation of China, Jiangsu Province’s Primary Research and Development Plan, the Fundamental Research Funds for the Central Universities, the Nanjing University PhD Student Zhujian Program and the Jiangsu Key Laboratory of Quantum Information Science and Technology at Nanjing University.
Paper data and sources
Original title: Stringent Constraints on Spin-Spin-Velocity-Dependent Exotic Interactions with a Levitated Magnet Force Sensor
Authors: Kenan Tian, Siwen Chen, Lei Wang et al.
Journal/Repository: arXiv
Status: Preprint, not yet peer-reviewed
First online: 2026-08-26
DOI: Not available
Original paper · Full text