Showing 1–29 of 29 results for author: Belov, D

HyperSurf: Quadruped Robot Leg Capable of Surface Recognition with GRU and Real-to-Sim Transferring

Authors: Sergei Satsevich, Yaroslav Savotin, Danil Belov, Elizaveta Pestova, Artem Erhov, Batyr Khabibullin, Artem Bazhenov, Vyacheslav Kovalev, Aleksey Fedoseev, Dzmitry Tsetserukou

Abstract: This paper introduces a system of data collection acceleration and real-to-sim transferring for surface recognition on a quadruped robot. The system features a mechanical single-leg setup capable of stepping on various easily interchangeable surfaces. Additionally, it incorporates a GRU-based Surface Recognition System, inspired by the system detailed in the Dog-Surf paper. This setup facilitates… ▽ More This paper introduces a system of data collection acceleration and real-to-sim transferring for surface recognition on a quadruped robot. The system features a mechanical single-leg setup capable of stepping on various easily interchangeable surfaces. Additionally, it incorporates a GRU-based Surface Recognition System, inspired by the system detailed in the Dog-Surf paper. This setup facilitates the expansion of dataset collection for model training, enabling data acquisition from hard-to-reach surfaces in laboratory conditions. Furthermore, it opens avenues for transferring surface properties from reality to simulation, thereby allowing the training of optimal gaits for legged robots in simulation environments using a pre-prepared library of digital twins of surfaces. Moreover, enhancements have been made to the GRU-based Surface Recognition System, allowing for the integration of data from both the quadruped robot and the single-leg setup. The dataset and code have been made publicly available. △ Less

Submitted 19 August, 2024; v1 submitted 22 July, 2024; originally announced July 2024.

Comments: IEEE SMC 2024

PartIR: Composing SPMD Partitioning Strategies for Machine Learning

Authors: Sami Alabed, Daniel Belov, Bart Chrzaszcz, Juliana Franco, Dominik Grewe, Dougal Maclaurin, James Molloy, Tom Natan, Tamara Norman, Xiaoyue Pan, Adam Paszke, Norman A. Rink, Michael Schaarschmidt, Timur Sitdikov, Agnieszka Swietlik, Dimitrios Vytiniotis, Joel Wee

Abstract: Training of modern large neural networks (NN) requires a combination of parallelization strategies encompassing data, model, or optimizer sharding. When strategies increase in complexity, it becomes necessary for partitioning tools to be 1) expressive, allowing the composition of simpler strategies, and 2) predictable to estimate performance analytically. We present PartIR, our design for a NN par… ▽ More Training of modern large neural networks (NN) requires a combination of parallelization strategies encompassing data, model, or optimizer sharding. When strategies increase in complexity, it becomes necessary for partitioning tools to be 1) expressive, allowing the composition of simpler strategies, and 2) predictable to estimate performance analytically. We present PartIR, our design for a NN partitioning system. PartIR is focused on an incremental approach to rewriting and is hardware-and-runtime agnostic. We present a simple but powerful API for composing sharding strategies and a simulator to validate them. The process is driven by high-level programmer-issued partitioning tactics, which can be both manual and automatic. Importantly, the tactics are specified separately from the model code, making them easy to change. We evaluate PartIR on several different models to demonstrate its predictability, expressibility, and ability to reach peak performance.. △ Less

Submitted 3 March, 2024; v1 submitted 20 January, 2024; originally announced January 2024.

Gamma Rhythm Analysis and Simulation Using Neuron Models

Authors: Evgeniia S. Sevasteeva, Sergei A. Plotnikov, Dmitry R. Belov

Abstract: Neural oscillations are electrical activities of the brain measurable at different frequencies. This paper studies the interaction between the fast and slow processes in the brain. We recorded signals intracranially from the simple Wistar rats, performed the signal processing, and computed the correlation between envelopes of the high-frequency gamma rhythm and a low-frequency ECoG signal. The ana… ▽ More Neural oscillations are electrical activities of the brain measurable at different frequencies. This paper studies the interaction between the fast and slow processes in the brain. We recorded signals intracranially from the simple Wistar rats, performed the signal processing, and computed the correlation between envelopes of the high-frequency gamma rhythm and a low-frequency ECoG signal. The analysis shows that the low-frequency signal (delta rhythm) modulates the gamma rhythm with a small time delay. Further, we used simple excitable neuron models, namely FitzHugh-Nagumo and Hindmarsh-Rose, to simulate the gamma rhythm. The low-frequency signal delta rhythm can be used as the input to affect the threshold and simulate gamma rhythm using these neuron models. △ Less

Submitted 10 May, 2022; originally announced May 2022.

Automap: Towards Ergonomic Automated Parallelism for ML Models

Authors: Michael Schaarschmidt, Dominik Grewe, Dimitrios Vytiniotis, Adam Paszke, Georg Stefan Schmid, Tamara Norman, James Molloy, Jonathan Godwin, Norman Alexander Rink, Vinod Nair, Dan Belov

Abstract: The rapid rise in demand for training large neural network architectures has brought into focus the need for partitioning strategies, for example by using data, model, or pipeline parallelism. Implementing these methods is increasingly supported through program primitives, but identifying efficient partitioning strategies requires expensive experimentation and expertise. We present the prototype o… ▽ More The rapid rise in demand for training large neural network architectures has brought into focus the need for partitioning strategies, for example by using data, model, or pipeline parallelism. Implementing these methods is increasingly supported through program primitives, but identifying efficient partitioning strategies requires expensive experimentation and expertise. We present the prototype of an automated partitioner that seamlessly integrates into existing compilers and existing user workflows. Our partitioner enables SPMD-style parallelism that encompasses data parallelism and parameter/activation sharding. Through a combination of inductive tactics and search in a platform-independent partitioning IR, automap can recover expert partitioning strategies such as Megatron sharding for transformer layers. △ Less

Submitted 6 December, 2021; originally announced December 2021.

Comments: Workshop on ML for Systems at NeurIPS 2021

Local Search for Policy Iteration in Continuous Control

Authors: Jost Tobias Springenberg, Nicolas Heess, Daniel Mankowitz, Josh Merel, Arunkumar Byravan, Abbas Abdolmaleki, Jackie Kay, Jonas Degrave, Julian Schrittwieser, Yuval Tassa, Jonas Buchli, Dan Belov, Martin Riedmiller

Abstract: We present an algorithm for local, regularized, policy improvement in reinforcement learning (RL) that allows us to formulate model-based and model-free variants in a single framework. Our algorithm can be interpreted as a natural extension of work on KL-regularized RL and introduces a form of tree search for continuous action spaces. We demonstrate that additional computation spent on model-based… ▽ More We present an algorithm for local, regularized, policy improvement in reinforcement learning (RL) that allows us to formulate model-based and model-free variants in a single framework. Our algorithm can be interpreted as a natural extension of work on KL-regularized RL and introduces a form of tree search for continuous action spaces. We demonstrate that additional computation spent on model-based policy improvement during learning can improve data efficiency, and confirm that model-based policy improvement during action selection can also be beneficial. Quantitatively, our algorithm improves data efficiency on several continuous control benchmarks (when a model is learned in parallel), and it provides significant improvements in wall-clock time in high-dimensional domains (when a ground truth model is available). The unified framework also helps us to better understand the space of model-based and model-free algorithms. In particular, we demonstrate that some benefits attributed to model-based RL can be obtained without a model, simply by utilizing more computation. △ Less

Submitted 12 October, 2020; originally announced October 2020.

V-MPO: On-Policy Maximum a Posteriori Policy Optimization for Discrete and Continuous Control

Authors: H. Francis Song, Abbas Abdolmaleki, Jost Tobias Springenberg, Aidan Clark, Hubert Soyer, Jack W. Rae, Seb Noury, Arun Ahuja, Siqi Liu, Dhruva Tirumala, Nicolas Heess, Dan Belov, Martin Riedmiller, Matthew M. Botvinick

Abstract: Some of the most successful applications of deep reinforcement learning to challenging domains in discrete and continuous control have used policy gradient methods in the on-policy setting. However, policy gradients can suffer from large variance that may limit performance, and in practice require carefully tuned entropy regularization to prevent policy collapse. As an alternative to policy gradie… ▽ More Some of the most successful applications of deep reinforcement learning to challenging domains in discrete and continuous control have used policy gradient methods in the on-policy setting. However, policy gradients can suffer from large variance that may limit performance, and in practice require carefully tuned entropy regularization to prevent policy collapse. As an alternative to policy gradient algorithms, we introduce V-MPO, an on-policy adaptation of Maximum a Posteriori Policy Optimization (MPO) that performs policy iteration based on a learned state-value function. We show that V-MPO surpasses previously reported scores for both the Atari-57 and DMLab-30 benchmark suites in the multi-task setting, and does so reliably without importance weighting, entropy regularization, or population-based tuning of hyperparameters. On individual DMLab and Atari levels, the proposed algorithm can achieve scores that are substantially higher than has previously been reported. V-MPO is also applicable to problems with high-dimensional, continuous action spaces, which we demonstrate in the context of learning to control simulated humanoids with 22 degrees of freedom from full state observations and 56 degrees of freedom from pixel observations, as well as example OpenAI Gym tasks where V-MPO achieves substantially higher asymptotic scores than previously reported. △ Less

Submitted 26 September, 2019; originally announced September 2019.

Comments: * equal contribution

TF-Replicator: Distributed Machine Learning for Researchers

Authors: Peter Buchlovsky, David Budden, Dominik Grewe, Chris Jones, John Aslanides, Frederic Besse, Andy Brock, Aidan Clark, Sergio Gómez Colmenarejo, Aedan Pope, Fabio Viola, Dan Belov

Abstract: We describe TF-Replicator, a framework for distributed machine learning designed for DeepMind researchers and implemented as an abstraction over TensorFlow. TF-Replicator simplifies writing data-parallel and model-parallel research code. The same models can be effortlessly deployed to different cluster architectures (i.e. one or many machines containing CPUs, GPUs or TPU accelerators) using synchr… ▽ More We describe TF-Replicator, a framework for distributed machine learning designed for DeepMind researchers and implemented as an abstraction over TensorFlow. TF-Replicator simplifies writing data-parallel and model-parallel research code. The same models can be effortlessly deployed to different cluster architectures (i.e. one or many machines containing CPUs, GPUs or TPU accelerators) using synchronous or asynchronous training regimes. To demonstrate the generality and scalability of TF-Replicator, we implement and benchmark three very different models: (1) A ResNet-50 for ImageNet classification, (2) a SN-GAN for class-conditional ImageNet image generation, and (3) a D4PG reinforcement learning agent for continuous control. Our results show strong scalability performance without demanding any distributed systems expertise of the user. The TF-Replicator programming model will be open-sourced as part of TensorFlow 2.0 (see https://github.com/tensorflow/community/pull/25). △ Less

Submitted 1 February, 2019; originally announced February 2019.

Relative Entropy Regularized Policy Iteration

Authors: Abbas Abdolmaleki, Jost Tobias Springenberg, Jonas Degrave, Steven Bohez, Yuval Tassa, Dan Belov, Nicolas Heess, Martin Riedmiller

Abstract: We present an off-policy actor-critic algorithm for Reinforcement Learning (RL) that combines ideas from gradient-free optimization via stochastic search with learned action-value function. The result is a simple procedure consisting of three steps: i) policy evaluation by estimating a parametric action-value function; ii) policy improvement via the estimation of a local non-parametric policy; and… ▽ More We present an off-policy actor-critic algorithm for Reinforcement Learning (RL) that combines ideas from gradient-free optimization via stochastic search with learned action-value function. The result is a simple procedure consisting of three steps: i) policy evaluation by estimating a parametric action-value function; ii) policy improvement via the estimation of a local non-parametric policy; and iii) generalization by fitting a parametric policy. Each step can be implemented in different ways, giving rise to several algorithm variants. Our algorithm draws on connections to existing literature on black-box optimization and 'RL as an inference' and it can be seen either as an extension of the Maximum a Posteriori Policy Optimisation algorithm (MPO) [Abdolmaleki et al., 2018a], or as an extension of Trust Region Covariance Matrix Adaptation Evolutionary Strategy (CMA-ES) [Abdolmaleki et al., 2017b; Hansen et al., 1997] to a policy iteration scheme. Our comparison on 31 continuous control tasks from parkour suite [Heess et al., 2017], DeepMind control suite [Tassa et al., 2018] and OpenAI Gym [Brockman et al., 2016] with diverse properties, limited amount of compute and a single set of hyperparameters, demonstrate the effectiveness of our method and the state of art results. Videos, summarizing results, can be found at goo.gl/HtvJKR . △ Less

Submitted 5 December, 2018; originally announced December 2018.

Parallel WaveNet: Fast High-Fidelity Speech Synthesis

Authors: Aaron van den Oord, Yazhe Li, Igor Babuschkin, Karen Simonyan, Oriol Vinyals, Koray Kavukcuoglu, George van den Driessche, Edward Lockhart, Luis C. Cobo, Florian Stimberg, Norman Casagrande, Dominik Grewe, Seb Noury, Sander Dieleman, Erich Elsen, Nal Kalchbrenner, Heiga Zen, Alex Graves, Helen King, Tom Walters, Dan Belov, Demis Hassabis

Abstract: The recently-developed WaveNet architecture is the current state of the art in realistic speech synthesis, consistently rated as more natural sounding for many different languages than any previous system. However, because WaveNet relies on sequential generation of one audio sample at a time, it is poorly suited to today's massively parallel computers, and therefore hard to deploy in a real-time p… ▽ More The recently-developed WaveNet architecture is the current state of the art in realistic speech synthesis, consistently rated as more natural sounding for many different languages than any previous system. However, because WaveNet relies on sequential generation of one audio sample at a time, it is poorly suited to today's massively parallel computers, and therefore hard to deploy in a real-time production setting. This paper introduces Probability Density Distillation, a new method for training a parallel feed-forward network from a trained WaveNet with no significant difference in quality. The resulting system is capable of generating high-fidelity speech samples at more than 20 times faster than real-time, and is deployed online by Google Assistant, including serving multiple English and Japanese voices. △ Less

Submitted 28 November, 2017; originally announced November 2017.

Parallel Multiscale Autoregressive Density Estimation

Authors: Scott Reed, Aäron van den Oord, Nal Kalchbrenner, Sergio Gómez Colmenarejo, Ziyu Wang, Dan Belov, Nando de Freitas

Abstract: PixelCNN achieves state-of-the-art results in density estimation for natural images. Although training is fast, inference is costly, requiring one network evaluation per pixel; O(N) for N pixels. This can be sped up by caching activations, but still involves generating each pixel sequentially. In this work, we propose a parallelized PixelCNN that allows more efficient inference by modeling certain… ▽ More PixelCNN achieves state-of-the-art results in density estimation for natural images. Although training is fast, inference is costly, requiring one network evaluation per pixel; O(N) for N pixels. This can be sped up by caching activations, but still involves generating each pixel sequentially. In this work, we propose a parallelized PixelCNN that allows more efficient inference by modeling certain pixel groups as conditionally independent. Our new PixelCNN model achieves competitive density estimation and orders of magnitude speedup - O(log N) sampling instead of O(N) - enabling the practical generation of 512x512 images. We evaluate the model on class-conditional image generation, text-to-image synthesis, and action-conditional video generation, showing that our model achieves the best results among non-pixel-autoregressive density models that allow efficient sampling. △ Less

Submitted 10 March, 2017; originally announced March 2017.

Quadratic Magnetoelectric Effect and Magnetic Field Induced Pyroelectric Effect in Multiferroic BaMnF4

Authors: A. K. Zvezdin, G. P. Vorob'ev, A. M. Kadomtseva, Yu. F. Popov, D. V. Belov, A. P. Pyatakov

Abstract: The experimental studies of magnetoelectric effects in pulse magnetic field up to 250 kOe and their theoretical analysis on the basis of magnetic symmetry consideration are carried out. It is shown that the nonvanishing components of quadratic magnetoelectric effect tensor corresponding to the electric polarization along b- and c-axes point out the triclinic distortion of the crystal symmetry. A… ▽ More The experimental studies of magnetoelectric effects in pulse magnetic field up to 250 kOe and their theoretical analysis on the basis of magnetic symmetry consideration are carried out. It is shown that the nonvanishing components of quadratic magnetoelectric effect tensor corresponding to the electric polarization along b- and c-axes point out the triclinic distortion of the crystal symmetry. Anomalous temperature dependence of magnetically induced polarization Pa(Hb) testifies to the magnetically induced pyroelectric effect. The torque curves measurements show the deflection of the spin orientation from the b-axis at 9 degrees of arc. △ Less

Submitted 8 November, 2008; originally announced November 2008.

Comments: 9 pages, 8 figures

T-duality, Gerbes and Loop Spaces

Authors: Dmitriy M. Belov, Chris M. Hull, Ruben Minasian

Abstract: We revisit sigma models on target spaces given by a principal torus fibration $X\to M$, and show how treating the 2-form B as a gerbe connection captures the gauging obstructions and the global constraints on the T-duality. We show that a gerbe connection on X, which is invariant with respect to the torus action, yields an affine double torus fibration Y over the base space M - the generalizatio… ▽ More We revisit sigma models on target spaces given by a principal torus fibration $X\to M$, and show how treating the 2-form B as a gerbe connection captures the gauging obstructions and the global constraints on the T-duality. We show that a gerbe connection on X, which is invariant with respect to the torus action, yields an affine double torus fibration Y over the base space M - the generalization of the correspondence space. We construct a symplectic form on the cotangent bundle to the loop space LY and study the relation of its symmetries to T-duality. We find that geometric T-duality is possible if and only if the torus symmetry is generated by Hamiltonian vector fields. Put differently, the obstruction to T-duality is the non-Hamiltonian action of the symmetry group. △ Less

Submitted 26 October, 2007; originally announced October 2007.

Comments: 32 pages, 5 figures,

Report number: Imperial/TP/07/DMB/01, SPhT-T07/134

Type II Actions from 11-Dimensional Chern-Simons Theories

Authors: Dmitriy M. Belov, Gregory W. Moore

Abstract: This paper continues the discussion of hep-th/0605038, applying the holographic formulation of self-dual theory to the Ramond-Ramond fields of type II supergravity. We formulate the RR partition function, in the presence of nontrivial H-fields, in terms of the wavefunction of an 11-dimensional Chern-Simons theory. Using the methods of hep-th/0605038 we show how to formulate an action principle f… ▽ More This paper continues the discussion of hep-th/0605038, applying the holographic formulation of self-dual theory to the Ramond-Ramond fields of type II supergravity. We formulate the RR partition function, in the presence of nontrivial H-fields, in terms of the wavefunction of an 11-dimensional Chern-Simons theory. Using the methods of hep-th/0605038 we show how to formulate an action principle for the RR fields of both type IIA and type IIB supergravity, in the presence of RR current. We find a new topological restriction on consistent backgrounds of type IIA supergravity, namely the fourth Wu class must have a lift to the H-twisted cohomology. △ Less

Submitted 2 November, 2006; originally announced November 2006.

Comments: 60 pages; 7 figures; LaTeX

Report number: Imperial/TP/06/DMB/02; RUNHETC-2006-28

Holographic Action for the Self-Dual Field

Authors: Dmitriy Belov, Gregory W. Moore

Abstract: We revisit the construction of self-dual field theory in 4l+2 dimensions using Chern-Simons theory in 4l+3 dimensions, building on the work of Witten. Careful quantization of the Chern-Simons theory reveals all the topological subtleties associated with the self-dual partition function, including the generalization of the choice of spin structure needed to define the theory. We write the partiti… ▽ More We revisit the construction of self-dual field theory in 4l+2 dimensions using Chern-Simons theory in 4l+3 dimensions, building on the work of Witten. Careful quantization of the Chern-Simons theory reveals all the topological subtleties associated with the self-dual partition function, including the generalization of the choice of spin structure needed to define the theory. We write the partition function for arbitrary torsion background charge, and in the presence of sources. We show how this approach leads to the formulation of an action principle for the self-dual field. △ Less

Submitted 3 May, 2006; originally announced May 2006.

Comments: 75 pages; 18 figures; LaTeX

Report number: NSF-KITP-05-120; Imperial/TP/06/DMB/01

Classification of abelian spin Chern-Simons theories

Authors: Dmitriy Belov, Gregory W. Moore

Abstract: We derive a simple classification of quantum spin Chern-Simons theories with gauge group T=U(1)^N. While the classical Chern-Simons theories are classified by an integral lattice the quantum theories are classified differently. Two quantum theories are equivalent if they have the same invariants on 3-manifolds with spin structure, or equivalently if they lead to equivalent projective representat… ▽ More We derive a simple classification of quantum spin Chern-Simons theories with gauge group T=U(1)^N. While the classical Chern-Simons theories are classified by an integral lattice the quantum theories are classified differently. Two quantum theories are equivalent if they have the same invariants on 3-manifolds with spin structure, or equivalently if they lead to equivalent projective representations of the modular group. We prove the quantum theory is completely determined by three invariants which can be constructed from the data in the classical action. We comment on implications for the classification of fractional quantum Hall fluids. △ Less

Submitted 25 May, 2005; originally announced May 2005.

Comments: 47 pages, 5 figures, LaTeX

Conformal blocks for AdS5 singletons

Authors: Dmitriy Belov, Gregory W. Moore

Abstract: We give a simple derivation of the conformal blocks of the singleton sector of compactifications of IIB string theory on spacetimes of the form X5 x Y5 with Y5 compact, while X5 has as conformal boundary an arbitrary 4-manifold M4. We retain the second-derivative terms in the action for the B,C fields and thus the analysis is not purely topological. The unit-normalized conformal blocks agree exa… ▽ More We give a simple derivation of the conformal blocks of the singleton sector of compactifications of IIB string theory on spacetimes of the form X5 x Y5 with Y5 compact, while X5 has as conformal boundary an arbitrary 4-manifold M4. We retain the second-derivative terms in the action for the B,C fields and thus the analysis is not purely topological. The unit-normalized conformal blocks agree exactly with the quantum partition function of the U(1) gauge theory on the conformal boundary. We reproduce the action of the magnetic translation group and the SL(2,Z) S-duality group obtained from the purely topological analysis of Witten. An interesting subtlety in the normalization of the IIB Chern-Simons phase is noted. △ Less

Submitted 15 December, 2004; originally announced December 2004.

Comments: 30pp. latex

Witten's Ghost Vertex Made Simple (bc and bosonized ghosts)

Authors: D. M. Belov

Abstract: First, we diagonalize the bc-ghost 3-string Neumann matrices using the technique described in hep-th/0304158. Their eigenvalues are in complete agreement with the previous authors. Second, we diagonalize the N-string gluing vertices for the bosonized ghost system. And third, we verify the descent and associativity relations for the combined bosonic matter+ghost gluing vertices. We find that in o… ▽ More First, we diagonalize the bc-ghost 3-string Neumann matrices using the technique described in hep-th/0304158. Their eigenvalues are in complete agreement with the previous authors. Second, we diagonalize the N-string gluing vertices for the bosonized ghost system. And third, we verify the descent and associativity relations for the combined bosonic matter+ghost gluing vertices. We find that in order for these relations to be true, the vertices must be normalized by the factor Z_N. Here Z_N is the partition function of the bosonic matter+ghost CFT on the gluing surface, which is the unit disc with the Neumann boundary conditions and the midpoint cone like singularity specifying by the angle excess π(N-2). △ Less

Submitted 28 September, 2003; v1 submitted 21 August, 2003; originally announced August 2003.

Comments: 23 pages, 2 figures, RevTeX style; ref. added

Report number: RUNHETC-2003-25

Journal ref: Phys.Rev. D69 (2004) 126001

Witten's Vertex Made Simple

Authors: D. Belov, C. Lovelace

Abstract: The infinite matrices in Witten's vertex are easy to diagonalize. It just requires some SL(2,R) lore plus a Watson-Sommerfeld transformation. We calculate the eigenvalues of all Neumann matrices for all scale dimensions s, both for matter and ghosts, including fractional s which we use to regulate the difficult s=0 limit. We find that s=1 eigenfunctions just acquire a p term, and x gets replaced… ▽ More The infinite matrices in Witten's vertex are easy to diagonalize. It just requires some SL(2,R) lore plus a Watson-Sommerfeld transformation. We calculate the eigenvalues of all Neumann matrices for all scale dimensions s, both for matter and ghosts, including fractional s which we use to regulate the difficult s=0 limit. We find that s=1 eigenfunctions just acquire a p term, and x gets replaced by the midpoint position. △ Less

Submitted 2 July, 2003; v1 submitted 17 April, 2003; originally announced April 2003.

Comments: 24 pages, 2 figures, RevTeX style, typos corrected

Report number: RUNHETC-2003-11

Journal ref: Phys.Rev. D68 (2003) 066003

Representation of small conformal algebra

Authors: D. M. Belov

Abstract: In hep-th/0202087 it was argued that the operator L_0 is bad defined in kappa-basis as a kernel operator. Indeed, we show that L_0 is a difference operator. We also find a representation of L_1 and L_{-1} in a class of difference operators. In hep-th/0202087 it was argued that the operator L_0 is bad defined in kappa-basis as a kernel operator. Indeed, we show that L_0 is a difference operator. We also find a representation of L_1 and L_{-1} in a class of difference operators. △ Less

Submitted 21 October, 2002; originally announced October 2002.

Comments: 12 pages, LaTeX 2e

On spectral density of Neumann matrices

Authors: D. M. Belov, A. Konechny

Abstract: In hep-th/0111281 the complete set of eigenvectors and eigenvalues of Neumann matrices was found. It was shown also that the spectral density contains a divergent constant piece that being regulated by truncation at level L equals (log L)/(2π). In this paper we find an exact analytic expression for the finite part of the spectral density. This function allows one to calculate finite parts of var… ▽ More In hep-th/0111281 the complete set of eigenvectors and eigenvalues of Neumann matrices was found. It was shown also that the spectral density contains a divergent constant piece that being regulated by truncation at level L equals (log L)/(2π). In this paper we find an exact analytic expression for the finite part of the spectral density. This function allows one to calculate finite parts of various determinants arising in string field theory computations. We put our result to some consistency checks. △ Less

Submitted 21 October, 2002; v1 submitted 17 October, 2002; originally announced October 2002.

Comments: 10 pages, 2 figures, LaTeX 2e; typos corrections

Journal ref: Phys.Lett. B558 (2003) 111-118

On Continuous Moyal Product Structure in String Field Theory

Authors: D. M. Belov, A. K. Konechny

Abstract: We consider a diagonalization of Witten's star product for a ghost system of arbitrary background charge and Grassmann parity. To this end we use a bosonized formulation of such systems and a spectral analysis of Neumann matrices. We further identify a continuous Moyal product structure for a combined ghosts+matter system. The normalization of multiplication kernel is discussed. We consider a diagonalization of Witten's star product for a ghost system of arbitrary background charge and Grassmann parity. To this end we use a bosonized formulation of such systems and a spectral analysis of Neumann matrices. We further identify a continuous Moyal product structure for a combined ghosts+matter system. The normalization of multiplication kernel is discussed. △ Less

Submitted 23 August, 2002; v1 submitted 18 July, 2002; originally announced July 2002.

Comments: 18+7 pages, 1 figure, typos corrections

Journal ref: JHEP 0210 (2002) 049

Diagonal Representation of Open String Star and Moyal Product

Authors: D. M. Belov

Abstract: We explicitly find the spectrum of the operators $M^{rs}$ and $\widetilde{M}^{rs}$, which specify the star-product in the matter and ghost sectors correspondingly. Further we derive the diagonal representation for the 3-string vertices. Using this representation we identify the appearing Moyal structures in the matter sector. In addition to the continuous non-commutativity parameter $θ(κ)$ found… ▽ More We explicitly find the spectrum of the operators $M^{rs}$ and $\widetilde{M}^{rs}$, which specify the star-product in the matter and ghost sectors correspondingly. Further we derive the diagonal representation for the 3-string vertices. Using this representation we identify the appearing Moyal structures in the matter sector. In addition to the continuous non-commutativity parameter $θ(κ)$ found in hep-th/0202087 we find the discrete non-commutativity parametrized by $θ_ξ$. △ Less

Submitted 9 May, 2002; v1 submitted 19 April, 2002; originally announced April 2002.

Comments: LaTeX 2e, 21+12 pages, 6 figures, typos corrections

Construction of the Vacuum String Field Theory on a non-BPS Brane

Authors: I. Ya. Aref'eva, D. M. Belov, A. A. Giryavets

Abstract: In the framework of the Sen conjectures a construction of vacuum superstring field theory on a non-BPS brane is discussed. A distinguished feature of this theory is a presence of a ghost kinetic operator mixing GSO+/- sectors. A candidate for such kinetic operator with zero cohomology is discussed. In the framework of the Sen conjectures a construction of vacuum superstring field theory on a non-BPS brane is discussed. A distinguished feature of this theory is a presence of a ghost kinetic operator mixing GSO+/- sectors. A candidate for such kinetic operator with zero cohomology is discussed. △ Less

Submitted 8 August, 2002; v1 submitted 24 January, 2002; originally announced January 2002.

Comments: expression for the pure ghost kinetic operator corrected, Comments added; 21 pages, 1 figure, LaTeX 2e

Journal ref: JHEP 0209 (2002) 050

Noncommutative Field Theories and (Super)String Field Theories

Authors: I. Ya. Aref'eva, D. M. Belov, A. A. Giryavets, A. S. Koshelev, P. B. Medvedev

Abstract: In this lecture notes we explain and discuss some ideas concerning noncommutative geometry in general, as well as noncommutative field theories and string field theories. We consider noncommutative quantum field theories emphasizing an issue of their renormalizability and the UV/IR mixing. Sen's conjectures on open string tachyon condensation and their application to the D-brane physics have led… ▽ More In this lecture notes we explain and discuss some ideas concerning noncommutative geometry in general, as well as noncommutative field theories and string field theories. We consider noncommutative quantum field theories emphasizing an issue of their renormalizability and the UV/IR mixing. Sen's conjectures on open string tachyon condensation and their application to the D-brane physics have led to wide investigations of the covariant string field theory proposed by Witten about 15 years ago. We review main ingredients of cubic (super)string field theories using various formulations: functional, operator, conformal and the half string formalisms. The main technical tools that are used to study conjectured D-brane decay into closed string vacuum through the tachyon condensation are presented. We describe also methods which are used to study the cubic open string field theory around the tachyon vacuum: construction of the sliver state, ``comma'' and matrix representations of vertices. △ Less

Submitted 18 January, 2002; v1 submitted 22 November, 2001; originally announced November 2001.

Comments: 160 pages, LaTeX, 29 EPS figures. Lectures given by I.Ya.Aref'eva at the Swieca Summer School, Brazil, January 2001; Summer School in Modern Mathematical Physics, Sokobanja, Yugoslavia, August 2001; Max Born Symposium, Karpacz, Poland, September, 2001; Workshop "Noncommutative Geometry, Strings and Renormalization", Leipzig, Germany, September 2001. Typos corrected, references added

Gauge Invariance and Tachyon Condensation in Cubic Superstring Field Theory

Authors: I. Ya. Aref'eva, D. M. Belov, A. S. Koshelev, P. B. Medvedev

Abstract: The gauge invariance of cubic open superstring field theory is considered in a framework of level truncation, and applications to the tachyon condensation problem are discussed. As it is known, in the bosonic case the Feynman-Siegel gauge is not universal within the level truncation method. We explore another gauge that is more suitable for calculation of the tachyon potential for fermionic stri… ▽ More The gauge invariance of cubic open superstring field theory is considered in a framework of level truncation, and applications to the tachyon condensation problem are discussed. As it is known, in the bosonic case the Feynman-Siegel gauge is not universal within the level truncation method. We explore another gauge that is more suitable for calculation of the tachyon potential for fermionic string at level (2,6). We show that this new gauge has no restrictions on the region of its validity at least at this level. △ Less

Submitted 4 September, 2001; v1 submitted 23 July, 2001; originally announced July 2001.

Comments: 21 pages, 2 figures, LaTeX 2e; references added, typos corrected

Journal ref: Nucl.Phys. B638 (2002) 21-40

Tachyon condensation in cubic superstring field theory

Authors: I. Ya. Aref'eva, D. M. Belov, A. S. Koshelev, P. B. Medvedev

Abstract: It has been conjectured that at the stationary point of the tachyon potential for the non-BPS D-brane or brane-anti-D-brane pair, the negative energy density cancels the brane tension. We study this conjecture using a cubic superstring field theory with insertion of a double-step inverse picture changing operator. We compute the tachyon potential at levels (1/2,1) and (2,6). In the first case we… ▽ More It has been conjectured that at the stationary point of the tachyon potential for the non-BPS D-brane or brane-anti-D-brane pair, the negative energy density cancels the brane tension. We study this conjecture using a cubic superstring field theory with insertion of a double-step inverse picture changing operator. We compute the tachyon potential at levels (1/2,1) and (2,6). In the first case we obtain that the value of the potential at the minimum is 97.5% of the non BPS D-brane tension. Using a special gauge in the second case we get 105.8% of the tension. △ Less

Submitted 23 July, 2001; v1 submitted 14 November, 2000; originally announced November 2000.

Comments: 19 pages, LaTeX, 3 figures. Eqs. (3.2), (3.3) and (4.6) are corrected, and new gauge fixing condition is used

Journal ref: Nucl.Phys.B638:3-20,2002

UV/IR Mixing for Noncommutative Complex Scalar Field Theory, II (Interaction with Gauge Fields)

Authors: I. Ya. Aref'eva, D. M. Belov, A. S. Koshelev, O. A. Rytchkov

Abstract: We consider noncommutative analogs of scalar electrodynamics and N=2 D=4 SUSY Yang-Mills theory. We show that one-loop renormalizability of noncommutative scalar electrodynamics requires the scalar potential to be an anticommutator squared. This form of the scalar potential differs from the one expected from the point of view of noncommutative gauge theories with extended SUSY containing a squar… ▽ More We consider noncommutative analogs of scalar electrodynamics and N=2 D=4 SUSY Yang-Mills theory. We show that one-loop renormalizability of noncommutative scalar electrodynamics requires the scalar potential to be an anticommutator squared. This form of the scalar potential differs from the one expected from the point of view of noncommutative gauge theories with extended SUSY containing a square of commutator. We show that fermion contributions restore the commutator in the scalar potential. This provides one-loop renormalizability of noncommutative N=2 SUSY gauge theory. We demonstrate a presence of non-integrable IR singularities in noncommutative scalar electrodynamics for general coupling constants. We find that for a special ratio of coupling constants these IR singularities vanish. Also we show that IR poles are absent in noncommutative N=2 SUSY gauge theory. △ Less

Submitted 20 March, 2000; originally announced March 2000.

Comments: 9 pages, 16 EPS figures

Report number: SMI-6-00

Journal ref: Nucl.Phys.Proc.Suppl. 102 (2001) 11-17

A Note on UV/IR for Noncommutative Complex Scalar Field

Authors: I. Ya. Aref'eva, D. M. Belov, A. S. Koshelev

Abstract: Noncommutative quantum field theory of a complex scalar field is considered. There is a two-coupling noncommutative analogue of U(1)-invariant quartic interaction $(φ^*φ)^2$, namely $Aφ^*\starφ\starφ^*\starφ+ Bφ^*\starφ^*\starφ\starφ$. For arbitrary values of $A$ and $B$ the model is nonrenormalizable. However, it is one-loop renormalizable in two special cases: B=0 and $A=B$. Furthermore, in th… ▽ More Noncommutative quantum field theory of a complex scalar field is considered. There is a two-coupling noncommutative analogue of U(1)-invariant quartic interaction $(φ^*φ)^2$, namely $Aφ^*\starφ\starφ^*\starφ+ Bφ^*\starφ^*\starφ\starφ$. For arbitrary values of $A$ and $B$ the model is nonrenormalizable. However, it is one-loop renormalizable in two special cases: B=0 and $A=B$. Furthermore, in the case B=0 the model does not suffer from IR divergencies at least at one-loop insertions level. △ Less

Submitted 31 January, 2000; originally announced January 2000.

Comments: 5 pages, 1 figure

Two-Loop Diagrams in Noncommutative $φ^4_4$ theory

Authors: I. Ya. Aref'eva, D. M. Belov, A. S. Koshelev

Abstract: Explicit two-loop calculations in noncommutative $φ^4_4$ theory are presented. It is shown that the model is two-loop renormalizable. Explicit two-loop calculations in noncommutative $φ^4_4$ theory are presented. It is shown that the model is two-loop renormalizable. △ Less

Submitted 14 December, 1999; v1 submitted 9 December, 1999; originally announced December 1999.

Comments: 11 pages, 9 figures, Table and references added, typos corrected

Report number: SMI-15-99

Journal ref: Phys.Lett.B476:431-436,2000