Advanced tokamak scenarios often feature weak or reversed magnetic shear configurations. In this study, the hybrid kinetic-MHD model implemented in the NIMROD code is used to investigate the effects of reversed magnetic shear on internal kink and fishbone mode in a circular shaped limiter tokamak. In the absence of energetic particles (EPs), the mode growth rate initially increases and then decreases as the magnetic shear changes from positive to negative, indicating stabilizing effects of the reversed magnetic shear on the internal kink mode. In the presence of EPs, when the reversed magnetic shear region is sufficiently narrow, the transition from internal kink/fishbone modes to double kink/fishbone modes takes place, and the stabilizing effects of the reversed magnetic shear can significantly dominate the destabilization of EPs. For non-resonant modes, the EP beta fraction $β_f$ for excitation increases with $q_{min}$, concurrent with progressively lower growth rates in non-resonant fishbone modes. When the equilibrium profile has an internal transport barrier (ITB), broader ITB widths suppress internal kink modes more effectively, whereas steeper temperature gradients strengthe
This review article will offer a qualitative overview of the strongly reversed shear profile for steady-state operation in tokamaks. For a steady-state reactor to be commercially viable, it is necessary to have a large bootstrap fraction. Currently, there appears great potential in an Advanced Tokamak (AT) regime, namely the hollow current profile (strongly reversed shear). This mode is characterized by high poloidal beta, broad current profiles, strong internal and edge pressure gradients, and relatively good magnetohydrodynamic (MHD) stability against Neoclassical Tearing Modes (NTMs) and ballooning modes. The n=1 and n=2 kink modes, resistive wall modes, and double tearing modes are of concern in the reversed shear profile, and avoidance and/or suppression of these modes is necessary. Although there is a relatively low net plasma current in the reversed shear, the regime appears to have excellent energy confinement properties due to the naturally occurring Internal Transport Barriers (ITBs) caused by the substantial bootstrap currents, and Edge Transport Barriers (ETBs), which can form from ELM-free H-Mode (QH-Mode), to form the Quiescent Double Barrier (QDBs). The reversed shea
The ripening dynamics in passive systems is governed by the theory of Lifshitz-Slyozov-Wagner (LSW). Here, we present an analog theory for reversed ripening in active systems. To derive the dynamic theory for the droplet size distribution, we consider a minimal ternary emulsion with one active reaction, leading to one conserved quantity. Even for cases where single droplets constantly grow, coupling many droplets via the conserved density in the far field leads to a self-organized reversal of ripening and, thus, a monodisperse emulsion. For late times, we find a scaling ansatz leading to the collapse of the rescaled size distributions, different from the LSW theory. This scaling behavior arises from a stable fixed point in the single droplet dynamics and may capture the late-time behavior of many active matter systems exhibiting reversed ripening.
The success of Transformer-based Language Models (LMs) stems from their attention mechanism. While this mechanism has been extensively studied in explainability research, particularly through the attention values obtained during the forward pass of LMs, the backward pass of attention has been largely overlooked. In this work, we study the mathematics of the backward pass of attention, revealing that it implicitly calculates an attention matrix we refer to as "Reversed Attention". We examine the properties of Reversed Attention and demonstrate its ability to elucidate the models' behavior and edit dynamics. In an experimental setup, we showcase the ability of Reversed Attention to directly alter the forward pass of attention, without modifying the model's weights, using a novel method called "attention patching". In addition to enhancing the comprehension of how LM configure attention layers during backpropagation, Reversed Attention maps contribute to a more interpretable backward pass.
We prove new results related to the digital reverse $\overleftarrow{n}$ of a positive integer $n$ in a fixed base $b$. First we show that for $b\geq 26000$, there exists infinitely many primes $p$ such that $\overleftarrow{p}$ is square-free. Further, we show that for $b\geq 2$ there are infinitely many palindromes (with $n=\overleftarrow{n}$) that are cube-free. We also give asymptotic expressions for the counting functions corresponding to these results. The main tools we use are recent bounds from the literature on reversed primes and palindromes in arithmetic progressions.
We investigate fixed points and cycle types of permutation polynomials and complete permutation polynomials arising from reversed Dickson polynomials of the first kind and second kind over $\mathbb{F}_p$. We also study the permutation behaviour of reversed Dickson polynomials of the first kind and second kind over $\mathbb{Z}_m$. Moreover, we prove two special cases of a conjecture on the permutation behaviour of reversed Dickson polynomials over $\mathbb{F}_p$.
Zeckendorf proved that every natural number $n$ can be expressed uniquely as a sum of non-consecutive Fibonacci numbers, called its Zeckendorf decomposition. Baird-Smith, Epstein, Flint, and Miller created the Zeckendorf game, a two-player game played on partitions of $n$ into Fibonacci numbers which always terminates at a Zeckendorf decomposition, and proved that Player 2 has a winning strategy for $n\geq 3$. Since their proof was non-constructive, other authors have studied the game to find a constructive winning strategy, and lacking success there turned to related problems. For example, Cheigh, Moura, Jeong, Duke, Milgrim, Miller, and Ngamlamai studied minimum and maximum game lengths and randomly played games. We explore a new direction and introduce the reversed Zeckendorf game, which starts at the ending state of the Zeckendorf game and flips all the moves, so the reversed game ends with all pieces in the first bin. We show that Player 1 has a winning strategy for $n = F_{i+1} + F_{i-2}$ and solve various modified games.
This work presents a discovery to advance the wisdom in a particular Succinct Data Structure: Wavelet Tree (Grossi, Gupta, and Vitter 2003). The discovery is first made by showing the feasibility of Reversed Indexes = Values: for integers within $[0,2^{N})$, there exists a Wavelet Tree that its compressed indexes can be equivalent to the Leibniz Binary system (Leibniz 1703), with only the bit reversal. Then we show how to strengthen the discovery by generalizing it into Reversed Indexes $\approx$ Values, by applying a longest common subsequence in bits and its patterns. Finally, we conjuncture potential implications of the above ideas by discussing its benefits, and modifications to the RAM model. The discovery reveals that: (1) the usability of Succinct Data Structure can be significantly expanded, by enabling Computation Directly on Compression; and (2) near-optimal lossless compression can still yield close connections with the Leibniz Binary System (Leibniz 1703), which breeds polymorphic functionalities within a single piece of the information. This work also provides an initial analysis of the benefits from the method (and potentially other extensions), and suggests potential
The reversed aging intensity function is defined as the ratio of the instantaneous reversed hazard rate to the baseline value of the reversed hazard rate. It analyzes the aging property quantitatively, the higher the reversed aging intensity, the weaker the tendency of aging. In this paper, a family of generalized reversed aging intensity functions is introduced and studied. Those functions depend on a real parameter. If the parameter is positive they characterize uniquely the distribution functions of univariate positive absolutely continuous random variables, in the opposite case they characterize families of distributions. Furthermore, the generalized reversed aging intensity orders are defined and studied. Finally, several numerical examples are given.
Lossless linear wave propagation is symmetric in time, a principle which can be used to create time reversed waves. Such waves are special 'pre-scattered' spatiotemporal fields, which propagate through a complex medium as if observing a scattering process in reverse, entering the medium as a complicated spatiotemporal field and arriving after propagation as a desired target field, such as a spatiotemporal focus. Time reversed waves have previously been demonstrated for relatively low frequency phenomena such as acoustics, water waves and microwaves. Many attempts have been made to extend these techniques into optics. However, the much higher frequencies of optics make for very different requirements. A fully time reversed wave is a volumetric field with arbitrary amplitude, phase and polarisation at every point in space and time. The creation of such fields has not previously been possible in optics. We demonstrate time reversed optical waves with a device capable of independently controlling all of light's classical degrees of freedom simultaneously. Such a class of ultrafast wavefront shaper is capable of generating a sequence of arbitrary 2D spatial/polarisation wavefronts at a
We study class of Lévy processes having distributions being indentifiable by moments. We define system of polynomial martingales ewline $\left\{ M_{n}(X_{t},t),\mathcal{F}_{\leq t}\right\} _{n\geq 1},$ where $% \mathcal{F}_{\leq t}$ is a suitable filtration defined below. We present several properties of these martingales. Among others we show that $% M_{1}(X_{t},t)/t$ is a reversed martingale as well as a harness. Main results of the paper concern the question if martingale say $M_{i}$ multiplied by suitable determinstic function $μ_{i}(t)$ is a reversed martingale. We show that for $n\geq 3$ $M_{n}(X_{t},t)$ is a reversed martingale (or orthogonal polynomial) only when the Lévy process in question is Gaussian (i.e. is a Wiener process). We study also a more general question if there are chances for a linear combination (with coefficients depending on $t)$ of martingales $M_{i},$ $i\allowbreak =\allowbreak 1,\ldots ,n$ to be reversed martingales. We analyze case $% n\allowbreak =\allowbreak 2$ in detail listing all possible cases.
Psychoacoustic studies have shown that locally-time reversed (LTR) speech, i.e., signal samples time-reversed within a short segment, can be accurately recognised by human listeners. This study addresses the question of how well a state-of-the-art automatic speech recognition (ASR) system would perform on LTR speech. The underlying objective is to explore the feasibility of deploying LTR speech in the training of end-to-end (E2E) ASR models, as an attempt to data augmentation for improving the recognition performance. The investigation starts with experiments to understand the effect of LTR speech on general-purpose ASR. LTR speech with reversed segment duration of 5 ms - 50 ms is rendered and evaluated. For ASR training data augmentation with LTR speech, training sets are created by combining natural speech with different partitions of LTR speech. The efficacy of data augmentation is confirmed by ASR results on speech corpora in various languages and speaking styles. ASR on LTR speech with reversed segment duration of 15 ms - 30 ms is found to have lower error rate than with other segment duration. Data augmentation with these LTR speech achieves satisfactory and consistent improv
On the basis of angular spectrum representation, the reversed propagation dynamics of Laguerre-Gaussian beam in left-handed materials (LHMs) is presented. We show that negative phase velocity gives rise to a reversed screw of wave-front, and ultimately leads to a reversed rotation of optical vortex. Furthermore, negative Gouy-phase shift causes an inverse spiral of Poynting vector. It is found that the Laguerre-Gaussian beam in LHMs will present the same propagation characteristics as the counterpart with opposite topological charges in regular right-handed materials (RHMs). The momentum conservation theorem insures that the tangential component of the wave momentum at the RHM-LHM boundary is conserved. It is shown that although the linear momentum reverses its direction, the angular momentum remains unchanged.
An analysis of experimental data from the RFX (Reversed Field eXperiment) reversed field pinch machine [L. Fellin, P. Kusstatscher and G. Rostagni, Fusion Eng. Des. {\bf 25}, 315 (1995)] is carried out to investigate the possible existence of deterministic chaos in the edge plasma region. The mathematical tools used include Lyapunov exponents, Kaplan-Yorke dimension, minimum embedding dimension estimates and nonlinear forecasting. The whole analysis agrees in ruling out the possibility of low-dimensional chaos: The dimension of the underlying dynamical system is estimated to be $ > 10$. From a critical re-reading of the literature it emerges that the findings of this work are likely to be common to all reversed field pinches.
The aim of this note is to connect a reversed form of the Gross logarithmic Sobolev inequality with the Gaussian maximum of Shannon's entropy power. There is thus a complete parallel with the well-known link between logarithmic Sobolev inequalities and their information theoretic counterparts. We moreover provide an elementary proof of the reversed Gross inequality via a two-point inequality and the Central Limit Theorem.
We show that solutions of free stochastic differential equations with regular drifts and diffusion coefficients, when considered backwards in time, still satisfy free SDEs for an explicit free Brownian motion and drift. We also study the stochastic integral part with respect to this reversed free Brownian motion of canonical closed martingales. We deduce that conjugate variables computed along a free Brownian motion, an example of such a reversed martingale appearing in the definition of non-microstates free entropy, are in the $L^2$ domain of corresponding free difference quotients for almost every time.
We prove new results on the additive theory of reversed primes $\overleftarrow{p}$; that is, primes $p$ which are written backwards in a fixed base $b\geq 2$. In particular, we study a variant of Goldbach's conjecture, looking at representations of integers as the sum of primes and reversed primes. We show that: (1) Every large odd integer is the sum of a prime and two reversed primes ($N=p_1+\overleftarrow{p_2}+\overleftarrow{p_3}$). (2) Every large odd integer is the sum of two primes and a reversed prime ($N=p_1+p_2+\overleftarrow{p_3}$). (3) Almost all even integers are the sum of a prime and a reversed prime ($N=p_1+\overleftarrow{p_2}$). (4) All large integers are the sum of a reversed prime and a square-free number ($N=\overleftarrow{p}+η$, $μ^2(η)=1$). To obtain our results, along with associated asymptotics, we apply the Hardy--Littlewood circle method and a novel refinement of the ``Zsiflaw--Legeis" theorem on the distribution of reversed primes in arithmetic progressions. Notably, our variant of the Zsiflaw--Legeis theorem does not require one to fix the digit length, unlike previous versions.
Recent advances in Large Language Models (LLMs) have opened new avenues for generating training-free text embeddings. However, the causal attention in decoder-only LLMs prevents earlier tokens from attending to future context, leading to biased contextualized representations. In this work, we propose Reverse prompting with Explicit One-word Limitation (ReverseEOL), a simple yet effective method for enhancing the representational capability of frozen LLMs. ReverseEOL augments the standard forward embedding with an additional reversed embedding derived from the reversed input text. Since reversing the input exposes each token to context inaccessible in the original order, the resulting reversed embedding effectively provides complementary information to the original one. As a result, combining the forward and reversed embeddings yields a richer final representation. Comprehensive experiments on STS and MTEB benchmarks demonstrate that ReverseEOL significantly improves the performance of existing training-free baselines across a broad range of LLMs with diverse architectures and scales. Extensive ablations and analyses further confirm the necessity of our reversal mechanism.
Motivated by applications in in-vivo DNA storage, we study codes for correcting duplications. A reverse-complement duplication of length $k$ is the insertion of the reversed and complemented copy of a substring of length $k$ adjacent to its original position, while a palindromic duplication only inserts the reversed copy without complementation. We first construct an explicit code with a single redundant symbol capable of correcting an arbitrary number of reverse-complement duplications (respectively, palindromic duplications), provided that all duplications have length $k \ge 3\lceil \log_q n \rceil$ and are disjoint. Next, we derive a Gilbert-Varshamov bound for codes that can correct a reverse-complement duplication (respectively, palindromic duplication) of arbitrary length, showing that the optimal redundancy is upper bounded by $2\log_q n + \log_q\log_q n + O(1)$. Finally, for $q \ge 4$, we present two explicit constructions of codes that can correct $t$ length-one reverse-complement duplications. The first construction achieves a redundancy of $2t\log_q n + O(\log_q\log_q n)$ with encoding complexity $O(n)$ and decoding complexity $O\big(n(\log_2 n)^4\big)$. The second const
Reversible Markov chains play a central role in stochastic modelling and in algorithms such as Markov chain Monte Carlo (MCMC). Motivated by the fundamental importance of reversibility in classical settings, this paper develops a theoretical framework for reversible imprecise Markov chains. We focus on their structural properties and their representation through joint distribution matrices. Adopting the strong independence interpretation, we reverse every precise chain compatible with a given imprecise Markov chain specification. Since the reversed ensemble generally cannot be encoded by the usual forward model defined by an imprecise initial distribution and a set of transition matrices, we introduce a symmetric representation based on credal sets of two-step joint distribution (or edge measure) matrices. This strictly more expressive framework naturally admits the reversal operation and reduces reversibility to simple matrix symmetry. Moreover, forward and reverse dynamics can be described simultaneously within a single closed convex set, providing a unified structural basis for the analysis of expectations of path-dependent functionals. We illustrate the theory with random walks