-
Analytic Properties of Trackable Weak Models
Authors:
Mark Chilenski,
George Cybenko,
Isaac Dekine,
Piyush Kumar,
Gil Raz
Abstract:
We present several new results on the feasibility of inferring the hidden states in strongly-connected trackable weak models. Here, a weak model is a directed graph in which each node is assigned a set of colors which may be emitted when that node is visited. A hypothesis is a node sequence which is consistent with a given color sequence. A weak model is said to be trackable if the worst case numb…
▽ More
We present several new results on the feasibility of inferring the hidden states in strongly-connected trackable weak models. Here, a weak model is a directed graph in which each node is assigned a set of colors which may be emitted when that node is visited. A hypothesis is a node sequence which is consistent with a given color sequence. A weak model is said to be trackable if the worst case number of such hypotheses grows as a polynomial in the sequence length. We show that the number of hypotheses in strongly-connected trackable models is bounded by a constant and give an expression for this constant. We also consider the problem of reconstructing which branch was taken at a node with same-colored out-neighbors, and show that it is always eventually possible to identify which branch was taken if the model is strongly connected and trackable. We illustrate these properties by assigning transition probabilities and employing standard tools for analyzing Markov chains. In addition, we present new results for the entropy rates of weak models according to whether they are trackable or not. These theorems indicate that the combination of trackability and strong connectivity dramatically simplifies the task of reconstructing which nodes were visited. This work has implications for any problem which can be described in terms of an agent traversing a colored graph, such as the reconstruction of hidden states in a hidden Markov model (HMM).
△ Less
Submitted 8 January, 2020;
originally announced January 2020.
-
Observability Properties of Colored Graphs
Authors:
Mark Chilenski,
George Cybenko,
Isaac Dekine,
Piyush Kumar,
Gil Raz
Abstract:
A colored graph is a directed graph in which nodes or edges have been assigned colors that are not necessarily unique. Observability problems in such graphs consider whether an agent observing the colors of edges or nodes traversed on a path in the graph can determine which node they are at currently or which nodes were visited earlier in the traversal. Previous research efforts have identified se…
▽ More
A colored graph is a directed graph in which nodes or edges have been assigned colors that are not necessarily unique. Observability problems in such graphs consider whether an agent observing the colors of edges or nodes traversed on a path in the graph can determine which node they are at currently or which nodes were visited earlier in the traversal. Previous research efforts have identified several different notions of observability as well as the associated properties of graphs for which those observability properties hold. This paper unifies the prior work into a common framework with several new results about relationships between those notions and associated graph properties. The new framework provides an intuitive way to reason about the attainable accuracy as a function of lag and time spent observing, and identifies simple modifications to improve the observability of a given graph. We show that one form of the graph modification problem is in NP-Complete. The intuition of the new framework is borne out with numerical experiments. This work has implications for problems that can be described in terms of an agent traversing a colored graph, including the reconstruction of hidden states in a hidden Markov model (HMM).
△ Less
Submitted 16 December, 2019; v1 submitted 9 November, 2018;
originally announced November 2018.
-
Control Flow Graph Modifications for Improved RF-Based Processor Tracking Performance
Authors:
Mark Chilenski,
George Cybenko,
Isaac Dekine,
Piyush Kumar,
Gil Raz
Abstract:
Many dedicated embedded processors do not have memory or computational resources to coexist with traditional (host-based) security solutions. As a result, there is interest in using out-of-band analog side-channel measurements and their analyses to accurately monitor and analyze expected program execution. In this paper, we describe an approach to this problem using externally observable multi-ban…
▽ More
Many dedicated embedded processors do not have memory or computational resources to coexist with traditional (host-based) security solutions. As a result, there is interest in using out-of-band analog side-channel measurements and their analyses to accurately monitor and analyze expected program execution. In this paper, we describe an approach to this problem using externally observable multi-band radio frequency (RF) measurements to make inferences about a program's execution. Because it is very difficult to identify individual instructions solely from their RF emissions, we compare RF measurements with the constrained execution logic of the program so that multiple RF measurements over time can effectively track program execution dynamically. In our approach, a program's execution is modeled by control flow graphs (CFG) and transitions between nodes of such graphs. We demonstrate that tracking performance can be improved through applications program modifications such as changing basic block transition properties and/or adding new basic blocks that are highly observable. In addition to demonstrating these principled approaches on some simple programs, we present initial results on the complexity and structure of real-world applications programs, namely gzip and md5sum, in this modeling framework.
△ Less
Submitted 31 October, 2019; v1 submitted 19 September, 2018;
originally announced September 2018.
-
Deep driven fMRI decoding of visual categories
Authors:
Michele Svanera,
Sergio Benini,
Gal Raz,
Talma Hendler,
Rainer Goebel,
Giancarlo Valente
Abstract:
Deep neural networks have been developed drawing inspiration from the brain visual pathway, implementing an end-to-end approach: from image data to video object classes. However building an fMRI decoder with the typical structure of Convolutional Neural Network (CNN), i.e. learning multiple level of representations, seems impractical due to lack of brain data. As a possible solution, this work pre…
▽ More
Deep neural networks have been developed drawing inspiration from the brain visual pathway, implementing an end-to-end approach: from image data to video object classes. However building an fMRI decoder with the typical structure of Convolutional Neural Network (CNN), i.e. learning multiple level of representations, seems impractical due to lack of brain data. As a possible solution, this work presents the first hybrid fMRI and deep features decoding approach: collected fMRI and deep learnt representations of video object classes are linked together by means of Kernel Canonical Correlation Analysis. In decoding, this allows exploiting the discriminatory power of CNN by relating the fMRI representation to the last layer of CNN (fc7). We show the effectiveness of embedding fMRI data onto a subspace related to deep features in distinguishing semantic visual categories based solely on brain imaging data.
△ Less
Submitted 9 January, 2017;
originally announced January 2017.
-
Constraining the Phase of B_s - \bar{B}_s Mixing
Authors:
Yuval Grossman,
Yosef Nir,
Guy Raz
Abstract:
New physics contributions to B_s-\bar{B}_s mixing can be parametrized by the size (r_s^2) and the phase (2θ_s) of the total mixing amplitude relative to the Standard Model amplitude. The phase has so far been unconstrained. We first use the D0 measurement of the semileptonic CP asymmetry A_SL to obtain the first constraint on the semileptonic CP asymmetry in B_s decays, A_SL^s=-0.008 \pm 0.011.…
▽ More
New physics contributions to B_s-\bar{B}_s mixing can be parametrized by the size (r_s^2) and the phase (2θ_s) of the total mixing amplitude relative to the Standard Model amplitude. The phase has so far been unconstrained. We first use the D0 measurement of the semileptonic CP asymmetry A_SL to obtain the first constraint on the semileptonic CP asymmetry in B_s decays, A_SL^s=-0.008 \pm 0.011. Then we combine recent measurements by the CDF and D0 collaborations - the mass difference (ΔM_s), the width difference (ΔΓ_s) and A_SL^s - to constrain 2θ_s. The errors on ΔΓ_s and A_SL^s should still be reduced to have a sensitive probe of the phase, yet the central values are such that the regions around 2θ_s ~ 3π/2 and, in particular, 2θ_s ~ π/2, are disfavored.
△ Less
Submitted 27 July, 2006; v1 submitted 2 May, 2006;
originally announced May 2006.
-
Four Generation CP Violation in B -> phi K^0, pi^0 K^0, eta' K^0 and Hadronic Uncertainties
Authors:
Wei-Shu Hou,
Makiko Nagashima,
Guy Raz,
Andrea Soddu
Abstract:
The fourth generation can give the correct trend of S_{phi K^0}, S_{pi^0 K^0} < sin2phi_1/beta, as indicated by data, and the effect, being largely leading order, is robust against hadronic uncertainties. The effect on S_{eta' K^0}, however, is diluted away by hadronic effects, and S_{eta' K^0} \simeq sin2phi_1/beta is expected. The near maximal arg[V^*_{t's}V_{t'b}] \lesssim 90^\circ that is ne…
▽ More
The fourth generation can give the correct trend of S_{phi K^0}, S_{pi^0 K^0} < sin2phi_1/beta, as indicated by data, and the effect, being largely leading order, is robust against hadronic uncertainties. The effect on S_{eta' K^0}, however, is diluted away by hadronic effects, and S_{eta' K^0} \simeq sin2phi_1/beta is expected. The near maximal arg[V^*_{t's}V_{t'b}] \lesssim 90^\circ that is needed could resolve the unequal direct CP violation seen in B -> K^+ pi^- and K^+ pi^0 modes, and is consistent with b ->s l^+l^- and B_s mixing constraints.
△ Less
Submitted 13 March, 2006;
originally announced March 2006.
-
Suppression of flavor symmetry breaking in B decay sum rules
Authors:
Michael Gronau,
Yuval Grossman,
Guy Raz,
Jonathan L. Rosner
Abstract:
While flavor symmetries are useful for studying hadronic B decays, symmetry relations for amplitudes and decay rates are usually violated by first order symmetry breaking corrections. We point out two cases in which first order symmetry breaking is suppressed by a small ratio of amplitudes: (1) An isospin sum rule for four $B\to Kπ$ decays, where isospin breaking is shown to be negligible. (2) A…
▽ More
While flavor symmetries are useful for studying hadronic B decays, symmetry relations for amplitudes and decay rates are usually violated by first order symmetry breaking corrections. We point out two cases in which first order symmetry breaking is suppressed by a small ratio of amplitudes: (1) An isospin sum rule for four $B\to Kπ$ decays, where isospin breaking is shown to be negligible. (2) An SU(3) sum rule for pairs of $B\to Kπ$ and $B\to Kη_8$, generalized to pairs of $B\to Kπ, B\to Kη$ and $B\to Kη'$.
△ Less
Submitted 29 January, 2006; v1 submitted 17 January, 2006;
originally announced January 2006.
-
Using SU(3) Flavor to Constrain the CP Asymmetries in $B \to PP,VP,VV$ Decays involving $b \to s$ Transitions
Authors:
Guy Raz
Abstract:
We use the approximate SU(3) flavor symmetry of the strong interaction to put bounds on the CP asymmetries in $b \to s$ decays of the $B^0_d$ and $B^+_u$ mesons. We extend the work of \cite{Grossman:2003qp} to include all relevant $B \to PP$, $B \to VP$ and $B \to VV$ decays. We obtain the strongest constraints from current data, and provide a list of SU(3) relations which can be used when futur…
▽ More
We use the approximate SU(3) flavor symmetry of the strong interaction to put bounds on the CP asymmetries in $b \to s$ decays of the $B^0_d$ and $B^+_u$ mesons. We extend the work of \cite{Grossman:2003qp} to include all relevant $B \to PP$, $B \to VP$ and $B \to VV$ decays. We obtain the strongest constraints from current data, and provide a list of SU(3) relations which can be used when future data is obtained.
△ Less
Submitted 13 September, 2005;
originally announced September 2005.
-
Using SU(3) Relations to bound the CP Asymmetries in $B \to KKK$ decays
Authors:
Guy Engelhard,
Guy Raz
Abstract:
We consider three body $Δs = 1$ $B \to f$ decays with $f=KKK$. The deviations of $-η_f S_f$ from $S_{ψK_S}$ and of $C_f$ from zero can be bounded using the approximate SU(3) flavor symmetry of the strong interactions and branching ratios of various $Δs = 0$ modes. We present the most promising SU(3) amplitude relations that can be used to obtain these bounds.
We consider three body $Δs = 1$ $B \to f$ decays with $f=KKK$. The deviations of $-η_f S_f$ from $S_{ψK_S}$ and of $C_f$ from zero can be bounded using the approximate SU(3) flavor symmetry of the strong interactions and branching ratios of various $Δs = 0$ modes. We present the most promising SU(3) amplitude relations that can be used to obtain these bounds.
△ Less
Submitted 3 August, 2005;
originally announced August 2005.
-
SU(3) Relations and the CP Asymmetry in $B \to K_S K_S K_S$
Authors:
Guy Engelhard,
Yosef Nir,
Guy Raz
Abstract:
The CP asymmetry in the $B \to K_S K_S K_S$ decay is being measured by the two B factories. A large deviation of the CP asymmetry $S_{K_S K_S K_S}$ from $-S_{ψK_S}$ and/or of $C_{K_S K_S K_S}$ from zero would imply new physics in $b \to s$ transitions. We try to put upper bounds on the Standard Model size of these deviations, using SU(3) flavor relations and experimental data on the branching ra…
▽ More
The CP asymmetry in the $B \to K_S K_S K_S$ decay is being measured by the two B factories. A large deviation of the CP asymmetry $S_{K_S K_S K_S}$ from $-S_{ψK_S}$ and/or of $C_{K_S K_S K_S}$ from zero would imply new physics in $b \to s$ transitions. We try to put upper bounds on the Standard Model size of these deviations, using SU(3) flavor relations and experimental data on the branching ratios of various decay modes that proceed via $b \to d$ transitions. We point out several subtleties that distinguish the case of three body final states from two body ones. We present several simple relations that can become useful once all relevant modes are measured accurately enough.
△ Less
Submitted 23 May, 2005;
originally announced May 2005.
-
The Pattern of CP Asymmetries in $b\to s$ Transitions
Authors:
Gerhard Buchalla,
Gudrun Hiller,
Yosef Nir,
Guy Raz
Abstract:
New CP violating physics in $b\to s$ transitions will modify the CP asymmetries in B decays into final CP eigenstates ($φK_S$, $η^\prime K_S$, $π^0 K_S$, $ωK_S$, $ρ^0 K_S$ and $ηK_S$) from their Standard Model values. In a model independent analysis, the pattern of deviations can be used to probe which Wilson coefficients get a significant contribution from the new physics. We demonstrate this i…
▽ More
New CP violating physics in $b\to s$ transitions will modify the CP asymmetries in B decays into final CP eigenstates ($φK_S$, $η^\prime K_S$, $π^0 K_S$, $ωK_S$, $ρ^0 K_S$ and $ηK_S$) from their Standard Model values. In a model independent analysis, the pattern of deviations can be used to probe which Wilson coefficients get a significant contribution from the new physics. We demonstrate this idea using several well-motivated models of new physics, and apply it to current data.
△ Less
Submitted 6 October, 2005; v1 submitted 16 March, 2005;
originally announced March 2005.
-
Light GUT Triplets and Yukawa Splitting
Authors:
Subhendu Rakshit,
Guy Raz,
Sourov Roy,
Yael Shadmi
Abstract:
Triplet-mediated proton decay in Grand Unified Theories (GUTs) is usually suppressed by arranging a large triplet mass. Here we explore instead a mechanism for suppressing the couplings of the triplets to the first and second generations compared to the Yukawa couplings, so that the triplets' mass can be below the GUT scale. This mechanism is based on a ``triplet symmetry'' in the context of pro…
▽ More
Triplet-mediated proton decay in Grand Unified Theories (GUTs) is usually suppressed by arranging a large triplet mass. Here we explore instead a mechanism for suppressing the couplings of the triplets to the first and second generations compared to the Yukawa couplings, so that the triplets' mass can be below the GUT scale. This mechanism is based on a ``triplet symmetry'' in the context of product-group GUTs. We study two possibilities. One, which requires the top Yukawa to arise from a non-renormalizable operator at the GUT scale, is that all triplet couplings to matter are negligible, so that the triplets can be at the weak scale. The second is that some triplet couplings, and in particular $T t b$ and $T \bar{t} \bar{l}$, are equal to the corresponding Yukawa couplings. This would give a distinct signature of grand unification if the triplets were sufficiently light. However, we derive a model-independent bound on the triplet mass in this case, which is at least 10$^6$GeV. Finally, we construct a GUT model based on Yukawa splitting, with the triplets at 10$^{14}$GeV, as required for coupling unification to work.
△ Less
Submitted 20 May, 2004; v1 submitted 29 September, 2003;
originally announced September 2003.
-
Time Variations in the Scale of Grand Unification
Authors:
Michael Dine,
Yosef Nir,
Guy Raz,
Tomer Volansky
Abstract:
We study the consequences of time variations in the scale of grand unification, $M_U$, when the Planck scale and the value of the unified coupling at the Planck scale are held fixed. We show that the relation between the variations of the low energy gauge couplings is highly model dependent. It is even possible, in principle, that the electromagnetic coupling $α$ varies, but the strong coupling…
▽ More
We study the consequences of time variations in the scale of grand unification, $M_U$, when the Planck scale and the value of the unified coupling at the Planck scale are held fixed. We show that the relation between the variations of the low energy gauge couplings is highly model dependent. It is even possible, in principle, that the electromagnetic coupling $α$ varies, but the strong coupling $α_3$ does not (to leading approximation). We investigate whether the interpretation of recent observations of quasar absorption lines in terms of time variation in $α$ can be accounted for by time variation in $M_U$. Our formalism can be applied to any scenario where a time variation in an intermediate scale induces, through threshold corrections, time variations in the effective low scale couplings.
△ Less
Submitted 8 October, 2002; v1 submitted 12 September, 2002;
originally announced September 2002.
-
Quark-Squark Alignment Revisited
Authors:
Yosef Nir,
Guy Raz
Abstract:
We re-examine the possibility that the solution to the supersymmetric flavor problem is related to small mixing angles in gaugino couplings induced by approximate horizontal Abelian symmetries. We prove that, for a large class of models, there is a single viable structure for the down quark mass matrix with four holomorphic zeros. Consequently, we are able to obtain both lower and upper bounds o…
▽ More
We re-examine the possibility that the solution to the supersymmetric flavor problem is related to small mixing angles in gaugino couplings induced by approximate horizontal Abelian symmetries. We prove that, for a large class of models, there is a single viable structure for the down quark mass matrix with four holomorphic zeros. Consequently, we are able to obtain both lower and upper bounds on the supersymmetric mixing angles and predict the contributions to various flavor changing neutral current processes. We find that the most likely signals for alignment are $Δm_D$ close to the present bound, significant CP violation in $D^0-\bar{D^0}$ mixing, and shifts of order a few percent in various CP asymmetries in $B^0$ and $B_s$ decays. In contrast, the modifications to radiative B decays, to $ε^\prime/ε$ and to $K\toπν\barν$ decays are small. We further investigate a new class of alignment models, where supersymmetric contributions to flavor changing processes are suppressed by both alignment and RGE-induced degeneracy.
△ Less
Submitted 6 June, 2002;
originally announced June 2002.
-
The mass insertion approximation without squark degeneracy
Authors:
Guy Raz
Abstract:
We study the applicability of the mass insertion approximation (MIA) for calculations of neutral meson mixing when squark masses are not degenerate and, in particular, in models of alignment. We show that the MIA can give results that are much better than an order of magnitude estimate as long as the masses are not strongly hierarchical. We argue that, in an effective two-squark framework, m_q=(…
▽ More
We study the applicability of the mass insertion approximation (MIA) for calculations of neutral meson mixing when squark masses are not degenerate and, in particular, in models of alignment. We show that the MIA can give results that are much better than an order of magnitude estimate as long as the masses are not strongly hierarchical. We argue that, in an effective two-squark framework, m_q=(m_1+m_2)/2 is the best choice for the MIA expansion point, rather than, for example, m_q^2=(m_1^2+m_2^2)/2.
△ Less
Submitted 28 May, 2002;
originally announced May 2002.
-
Interpreting experimental bounds on D^0 - \bar{D^0} mixing in the presence of CP violation
Authors:
Guy Raz
Abstract:
We analyse the most recent experimental data regarding D^0 - \bar{D^0} mixing, allowing for CP violation. We focus on the dispersive part of the mixing amplitude, M^D_{12}, which is sensitive to new physics contributions. We obtain a constraint on the mixing amplitude: |M^D_{12}| < 6.2\times 10^{-11} MeV at 95% C.L. . This constraint is weaker by a factor of about three than the one which is obt…
▽ More
We analyse the most recent experimental data regarding D^0 - \bar{D^0} mixing, allowing for CP violation. We focus on the dispersive part of the mixing amplitude, M^D_{12}, which is sensitive to new physics contributions. We obtain a constraint on the mixing amplitude: |M^D_{12}| < 6.2\times 10^{-11} MeV at 95% C.L. . This constraint is weaker by a factor of about three than the one which is obtained when no CP violation is assumed.
△ Less
Submitted 27 May, 2002; v1 submitted 12 May, 2002;
originally announced May 2002.
