VAC: A Multifaceted Research Term

• VAC is a context-sensitive abbreviation with distinct definitions across fields, such as vergence-accommodation conflict in VR and variational approaches in molecular dynamics.
• It underpins practical methodologies in video action counting, generative modeling (VAC+GAN and VAC-CNN), and vehicular admission control for urban traffic systems.
• Disambiguation is critical as similar notations (e.g., vacancy formation energies, vacuum energy density) appear in materials science and fundamental physics.

VAC is a domain-dependent abbreviation rather than a single standardized concept. In recent arXiv literature, it denotes at least five distinct technical objects: the vergence-accommodation conflict in virtual reality, the variational approach to conformational dynamics in molecular kinetics, video action counting in computer vision, vehicular admission control in urban traffic control, and the Versatile Auxiliary Classifier in generative adversarial learning. A recurrent source of ambiguity is that the string “vac” also appears as a subscript or model label in unrelated settings—most notably vacancy formation energies $E_{\rm vac}$, vacuum energy density $\rho_{\rm vac}$, the vacuum gauge field $A_{\sf{vac}}$, and vac=1/vac=0 model variants—without functioning as the acronym “VAC” (Wang et al., 29 May 2025, Lorpaiboon et al., 2020, Wang et al., 2024, Ramp et al., 2 Jan 2026, Bazrafkan et al., 2018, Choudhary et al., 2022).

1. Scope and disambiguation

In current research usage, “VAC” is best understood as a context-sensitive label whose meaning is fixed by disciplinary conventions rather than by a universal expansion. The table summarizes the principal senses represented in recent arXiv work.

Research area	Expansion or usage	Representative papers
Virtual reality	Vergence-accommodation conflict	(Wang et al., 29 May 2025)
Molecular dynamics	Variational approach to conformational dynamics; Integrated VAC	(Lorpaiboon et al., 2020, Webber et al., 2020)
Computer vision	Video Action Counting; Irregular Video Action Counting (IVAC)	(Wang et al., 2024)
Generative modeling	Versatile Auxiliary Classifier in VAC+GAN	(Bazrafkan et al., 2018, Bazrafkan et al., 2018)
Visual analytics	VAC-CNN = Visual Analytics for Comparing CNNs	(Xuan et al., 2021)
Traffic systems	Vehicular admission control	(Ramp et al., 2 Jan 2026)
Materials and fundamental physics	$E_{\rm vac}$, $\rho_{\rm vac}$, $A_{\sf{vac}}$, vac=1/vac=0	(Choudhary et al., 2022, Peracaula, 2021, Popov, 2024, Singh et al., 8 Jun 2026)

A common misconception is that “VAC” in one literature transfers directly to another. The cited papers show the opposite: each usage carries its own mathematical objects, observables, and validation criteria. Another frequent confusion is between acronymic and subscripted uses. In $E_{\rm vac}$ and $\rho_{\rm vac}$, “vac” denotes “vacancy” or “vacuum” as a suffix-like label, not the standalone acronym VAC (Choudhary et al., 2022, Peracaula, 2021).

2. VAC as vergence-accommodation conflict in virtual reality

In virtual reality research, VAC denotes the vergence-accommodation conflict, the mismatch created because vergence can vary with virtual object distance while accommodation remains fixed at the screen distance of the head-mounted display. In natural vision, vergence and accommodation are congruent; in VR HMDs, their decoupling produces discomfort, depth misperception, and degraded visually guided actions such as reaching and pointing (Wang et al., 29 May 2025).

The 2025 geometrical analysis models VAC as a constant vergence angle offset: $\hat{\phi}=\phi+\beta_{\rm offset},$ with effective target angle

$\hat{\tau}=(\phi+\beta_{\rm offset})-\delta,$

and perceived target distance

$\rho_{\rm vac}$0

The corresponding depth error is written as

$\rho_{\rm vac}$1

For the HTC VIVE Pro, the fitted vergence offset was reported as approximately $\rho_{\rm vac}$2–$\rho_{\rm vac}$3, corresponding to about $\rho_{\rm vac}$4 cm at typical viewing distances (Wang et al., 29 May 2025).

The first experiment used a 3D pointing task with Online Guidance and Feedforward conditions. VAC primarily affected the Online Guidance condition, where systematic undershooting increased with distance; the model accounted for $\rho_{\rm vac}$5 of variance in test data for online guidance with $\rho_{\rm vac}$6. Feedforward movements were not sensitive to the VAC-induced offset, which the paper interprets as evidence that the error source lies in disparity matching during real-time visual control rather than in memory for target distance (Wang et al., 29 May 2025).

The same work introduced a software correction implemented as a vertex shader in Unity. For each rendered point, the transformation computes

$\rho_{\rm vac}$7

The transformation modifies only the $\rho_{\rm vac}$8 coordinate, leaving $\rho_{\rm vac}$9 and $A_{\sf{vac}}$0 unchanged. The reported advantages were that it required no eye tracking, had low computational overhead, and was deployable on existing HMDs (Wang et al., 29 May 2025).

A second experiment validated this correction. With online feedback, the Transformed condition improved pointing accuracy by approximately 30\%, corresponding to a 0.73 cm improvement, for the majority of participants. The effect was neutral or slightly negative for a minority, indicating inter-individual variability. The paper also notes residual distance-dependent undershooting, suggesting that a constant offset may not completely eliminate all depth errors (Wang et al., 29 May 2025).

3. VAC as variational approach to conformational dynamics

In molecular simulation and dynamical systems, VAC denotes the variational approach to conformational dynamics, a spectral estimation framework for identifying slowly decorrelating modes of a stationary, ergodic continuous-time Markov process. The transition operator is

$A_{\sf{vac}}$1

and VAC seeks functions with large normalized time-autocorrelation,

$A_{\sf{vac}}$2

Under the usual self-adjoint setting, the target objects are eigenfunctions of the transition operator, and practical implementations reduce to generalized eigenvalue problems in linear bases or to nonlinear optimization with flexible parameterizations (Lorpaiboon et al., 2020).

The central practical issue is the lag time $A_{\sf{vac}}$3. The 2020 error-analysis paper decomposes total error into approximation error and estimation error, showing that short lag times favor lower approximation bias whereas long lag times amplify finite-sample error because the relevant spectral gaps shrink (Webber et al., 2020). The paper defines a VAC condition number,

$A_{\sf{vac}}$4

to quantify ill-conditioning for a target eigenspace. Estimation error scales inversely with the minimum spectral gap and asymptotically as $A_{\sf{vac}}$5 with data length $A_{\sf{vac}}$6 (Webber et al., 2020).

The Integrated VAC (IVAC) extension addresses lag-time sensitivity by integrating correlations over a window of lag times rather than optimizing at a single $A_{\sf{vac}}$7. The continuous objective is

$A_{\sf{vac}}$8

and, in the linear discrete setting, the method solves

$A_{\sf{vac}}$9

This integrated formulation was proposed specifically to make results more robust and reproducible than standard single-lag VAC (Lorpaiboon et al., 2020).

The empirical demonstrations on alanine dipeptide and villin headpiece showed that IVAC reduced sensitivity to lag-time choice and mitigated overfitting in neural-network parameterizations. The paper reports that standard VAC can produce highly variable implied timescales and periodic artifacts at the training lag time, whereas IVAC yields smoother spectra and more consistent eigenfunction estimates across runs (Lorpaiboon et al., 2020). A plausible implication is that IVAC is best viewed not as a different target operator, but as a regularized estimator for essentially the same slow dynamical subspaces.

4. VAC as video action counting

In computer vision, VAC commonly denotes Video Action Counting, the task of estimating how many repetitions of an action occur in a video. The 2024 paper on Irregular Video Action Counting (IVAC) argues that prior VAC methods under-modeled interruptions and variable cycle duration, and it formalizes irregular repetition through two priors: Inter-cycle Consistency and Cycle-interval Inconsistency (Wang et al., 2024).

The end-to-end formulation treats counting as regression,

$E_{\rm vac}$0

with segment embeddings constructed by mean pooling framewise representations over cycle and interval segments: $E_{\rm vac}$1 The model introduces a pull-push loss,

$E_{\rm vac}$2

where, in practice, $E_{\rm vac}$3, $E_{\rm vac}$4, and $E_{\rm vac}$5 are all set to $E_{\rm vac}$6. The pull term

$E_{\rm vac}$7

encourages coherence among cycles, while the push term

$E_{\rm vac}$8

separates cycle and interval features (Wang et al., 2024).

Performance is reported using MAE and OBO. On RepCount-A, IVAC-$E_{\rm vac}$9 achieved MAE $\rho_{\rm vac}$0 and OBO $\rho_{\rm vac}$1, improving on TransRAC at MAE $\rho_{\rm vac}$2 and OBO $\rho_{\rm vac}$3. In zero-shot evaluation, the same model obtained MAE $\rho_{\rm vac}$4 / OBO $\rho_{\rm vac}$5 on UCFRep and MAE $\rho_{\rm vac}$6 / OBO $\rho_{\rm vac}$7 on Countix, outperforming TransRAC on both datasets without dataset-specific fine-tuning (Wang et al., 2024).

The ablations emphasize that the push loss is particularly important, that contrastive and triplet losses underperform the proposed structure, and that whole-cycle consistency is more effective than finer phase partitioning. The paper also introduces Random Count Augmentation (RCA) to address unbalanced cycle counts (Wang et al., 2024). This suggests that, in the VAC setting, explicit priors on irregular repetition are more consequential than purely periodicity-driven representations.

5. VAC-derived acronyms in machine learning: VAC+GAN and VAC-CNN

A different machine-learning usage appears in VAC+GAN, short for Versatile Auxiliary Classifier with Generative Adversarial Network. The architecture places a classifier in parallel with the discriminator rather than embedding classification inside the discriminator. Generated samples are passed to the classifier, and the classification loss is backpropagated through both the classifier and the generator, while the discriminator is trained independently (Bazrafkan et al., 2018, Bazrafkan et al., 2018).

For the multi-class formulation, the reported losses are

$\rho_{\rm vac}$8

$\rho_{\rm vac}$9

The theoretical result is that minimizing the categorical cross-entropy increases the Jensen-Shannon divergence among class-conditional generated distributions: $A_{\sf{vac}}$0 The papers argue that this decoupled design is applicable to any GAN implementation, in contrast to ACGAN-style formulations that entangle classification with discriminator structure (Bazrafkan et al., 2018).

Empirically, the multi-class paper reports that on CIFAR-10, VAC+GAN achieved 74.49\% classification accuracy versus 71.89\% for ACGAN. On MNIST, the method was compared to CGAN, CDCGAN, and ACGAN, with the stated conclusion that VAC+GAN produced clearer class separation while preserving architectural flexibility (Bazrafkan et al., 2018). The earlier binary-setting paper used BEGAN on CelebA and reported higher diversity and lower inter-class similarity than CGAN-style conditioning, with $A_{\sf{vac}}$1 and $A_{\sf{vac}}$2 in the generator loss (Bazrafkan et al., 2018).

A separate derived acronym is VAC-CNN, expanded as Visual Analytics for Comparing CNNs. This is a web-based system for comparative inspection of two or more CNN models, explicitly designed to scale to tens of models. Its workflow consists of Information Overview, Task Customization, and Model Investigation and Comparison, implemented through five coordinated views: Overall Information View, Distribution Graph View, Task Selection Sidebar, Visual Explanation View, and Supplemental View (Xuan et al., 2021).

VAC-CNN integrates quantitative metrics such as accuracy and computational complexity with qualitative explanations including Grad-CAM, BBMP, Grad-CAM++, Smooth Grad-CAM++, and Score-CAM. It also computes class-distance matrices and explanation-similarity matrices using measures such as L1, MSE, SSIM, and perceptual hash (Xuan et al., 2021). In a preliminary evaluation with 12 graduate students, the system received median scores of $A_{\sf{vac}}$3 for ease of use and $A_{\sf{vac}}$4 for helpfulness, and all users were able to extract insights on model behavior and performance (Xuan et al., 2021).

6. VAC as vehicular admission control

In transportation systems, VAC denotes vehicular admission control, a family of strategies that regulate vehicle entry into urban traffic networks. The 2026 study examines decentralized VAC in large-scale, heterogeneous networks with nonlinear region dynamics, concave macroscopic fundamental diagrams, and bounded uncertainty between flow, density, and speed (Ramp et al., 2 Jan 2026).

The network dynamics for region $A_{\sf{vac}}$5 are written as

$A_{\sf{vac}}$6

where $A_{\sf{vac}}$7 is admitted demand, and interregional flow is modeled as

$A_{\sf{vac}}$8

The uncertainty term $A_{\sf{vac}}$9 is assumed Lipschitz continuous, bounded, and compatible with concavity and unimodality of the total flow (Ramp et al., 2 Jan 2026).

Each local VAC law is described by a general nonlinear state-space controller,

$E_{\rm vac}$0

driven solely by local density feedback. The analysis imposes local input strict passivity through a storage function and derives a distributed stability condition involving the controller passivity gain $E_{\rm vac}$1, the Lipschitz constants of the nominal MFD and uncertainty, and coupling terms from neighboring regions. The paper emphasizes that these conditions are scalable, locally verifiable, and valid for arbitrary connected network layouts (Ramp et al., 2 Jan 2026).

Several controller classes are given explicitly: Proportional Control, Proportional + Static Nonlinearity, First-Order Dynamic Regulation, Second-Order Linear Dynamics, and an Integrator-Passive Scheme for feasible setpoint tracking (Ramp et al., 2 Jan 2026). Numerical simulations on a 6-region system included complete temporary loss of control and constant random driver non-compliance; in both cases the network returned to desired operating points without gridlock. The paper further states that similar behavior was obtained on a 20-region network (Ramp et al., 2 Jan 2026). This establishes VAC, in the traffic sense, as a control-theoretic object centered on robustness under modeling uncertainty rather than on optimization alone.

7. Related but non-equivalent “vac” notations in materials science and fundamental physics

Outside acronymic usage, “vac” appears extensively as a notation for vacancy or vacuum. In materials informatics, the vacancy formation energy is

$E_{\rm vac}$2

A 2022 study showed that ALIGNN, trained only on perfect structures, can predict neutral vacancy formation energies for defective structures without defect data or additional retraining. The benchmark dataset contained 508 DFT vacancy formation energies, the raw MAE was 1.51 eV, and a heuristic 1.3 eV scissor shift reduced MAE to 1.0 eV. The model was then used to predict 192,494 vacancy energies for 55,723 materials in JARVIS-DFT (Choudhary et al., 2022). Here, “vac” labels a defect observable rather than the acronym VAC.

In cosmology and quantum field theory, $E_{\rm vac}$3 denotes vacuum energy density. Running-vacuum studies derive

$E_{\rm vac}$4

arguing that adiabatic regularization in FLRW spacetime yields the RVM structure and removes dangerous quartic mass terms $E_{\rm vac}$5 from the renormalized expression (Peracaula, 2021, Moreno-Pulido et al., 2021). A related equation-of-state analysis states that the running vacuum is close to $E_{\rm vac}$6 during inflation, tracks radiation with $E_{\rm vac}$7, tracks matter with $E_{\rm vac}$8, and becomes mildly quintessence-like, $E_{\rm vac}$9, in the late universe (Moreno-Pulido et al., 2022). The 2025 inflation paper further argues that an $\rho_{\rm vac}$0 term can drive an inflationary phase with approximately constant $\rho_{\rm vac}$1 and no supplementary inflaton field (Peracaula et al., 2 Mar 2025).

A different flat-spacetime proposal defines $\rho_{\rm vac}$2 thermodynamically through the high-temperature expansion of the free energy and obtains the general form

$\rho_{\rm vac}$3

with the sign depending on the theory; for lattice QCD with two light quarks and a strange quark, the estimate reported is

$\rho_{\rm vac}$4

(LeClair, 2024). Other QFT approaches include the interacting-theory proposal

$\rho_{\rm vac}$5

for $\rho_{\rm vac}$6-type settings (LeClair, 2023), and a form-factor-bootstrap prescription yielding universal scaling $\rho_{\rm vac}$7 and a speculative “zeron” interpretation, with a massive Majorana neutrino named as a strong candidate (LeClair, 2024).

The notation also appears in geometric and QCD-adjacent contexts. The vacuum Abelian gauge field proposal defines

$\rho_{\rm vac}$8

and interprets confinement of quarks and gluons as arising from an always-on interaction with this vacuum field (Popov, 2024). In finite-temperature QCD effective modeling, vac=1 and vac=0 distinguish inclusion or omission of the fermion vacuum term in the PCQMF model; at $\rho_{\rm vac}$9 MeV, the paper reports $\hat{\phi}=\phi+\beta_{\rm offset},$0 MeV and $\hat{\phi}=\phi+\beta_{\rm offset},$1 MeV for vac=1, versus $\hat{\phi}=\phi+\beta_{\rm offset},$2 MeV and $\hat{\phi}=\phi+\beta_{\rm offset},$3 MeV for vac=0, together with marked differences in higher-order susceptibilities and cumulant ratios (Singh et al., 8 Jun 2026).

Taken together, these usages show that “VAC” and “vac” occupy a broad semantic field spanning perception, kinetics, counting, generative modeling, traffic control, defect energetics, and vacuum physics. Interpretation is therefore inseparable from disciplinary context.