Memory Mode Research Articles

Two distinct scene processing networks connecting vision and memory

Research on visual scene understanding has identified a number of regions involved in processing natural scenes, but has lacked a unifying framework for understanding how these different regions are organized and interact. We propose a new organizational principle, in which scene processing relies on two distinct networks at the edge of visual cortex. The first network consists of the Transverse Occipital Sulcus (TOS, or the Occipital Place Area) and the posterior portion of the Parahippocampal Place Area (PPA). These regions have a well-defined retinotopic organization and do not show strong memory or context effects, suggesting that this network primarily processes visual features from the current view of a scene. The second network consists of the caudal Inferior Parietal Lobule (cIPL), Retrosplenial Cortex (RSC), and the anterior portion of the PPA. These regions are involved in a wide range of both visual and non-visual tasks involving episodic memory, navigation, imagination, and default mode processing, and connect information about a current scene view with a much broader temporal and spatial context. We provide evidence for this division from a diverse set of sources. Using a data-driven approach to parcellate resting-state fMRI data, we identify coherent functional regions corresponding to scene-processing areas. We then show that a network clustering analysis separates these scene-related regions into two adjacent networks, which exhibit sharp changes in connectivity properties across their narrow border. Additionally, we argue that the cIPL has been previously overlooked as a critical region for full scene understanding, based on a meta-analysis of previous functional studies as well as diffusion tractography results showing that cIPL is well-positioned to connect visual cortex with many other cortical systems. This new framework for understanding the neural substrates of scene processing bridges results from many lines of research, and makes specific predictions about functional properties of these regions. Meeting abstract presented at VSS 2015

High‐Performance n‐Type Electrical Memory and Morphology‐Induced Memory‐Mode Tuning of a Well‐Defined Brush Polymer Bearing Perylene Diimide Moieties

Poly(N‐(1‐hexylheptyl)‐N′‐(12‐oxydodecyl)perylene‐3,4,9,10‐tetracarboxyldiimide acrylate) (PAcPDI), a perylene diimide (PDI) containing brush polymer, is synthesized, revealing good solubility in organic solvents, excellent thermal stability up to around 340 °C, and two melting transitions over 130−220 °C. The self‐assembly and n‐type memory characteristics of PAcPDI in nanoscale thin films are quantitatively investigated. As‐cast films of PAcPDI are completely amorphous and the PDI units nevertheless formed π–π stacks favorably. However, the PAcPDI molecules can self‐assemble via thermal annealing, developing a well‐ordered horizontal lamellar structure with monomorphic or polymorphic monoclinic PDI crystals. The formation of monomorphic or polymorphic monoclinic crystals is attributed to various π–π stack modes of the PDI units, and is shown to be dependent on the film thickness. The differences in the thin film morphologies are directly reflected into the electrical memory behavior. The thermally annealed films demonstrate high‐performance n‐type unipolar volatile memory behavior within the thickness range of 12−31 nm. The as‐cast films show n‐type unipolar nonvolatile or volatile memory behavior in the range of 12−53 nm. The memory mode of PAcPDI films can be tuned by changing either the morphology or the film thickness.

Replication in interaction and working memory research: Révész (2012) and Goo (2012)

This paper argues for the replication of two studies, both of which consider feedback and working memory. In the first part of this paper, we discuss the role of interaction-based research and working memory research in second language acquisition research. We then describe two studies that have unified these two areas in recent published articles and discuss ways that replication can further our knowledge in how working memory capacity can impact the benefits of feedback. We emphasize the importance of replication in a number of different areas, including grammatical complexity of the target structure and the nature of working memory tests, in scoring, actual tests, and modes of delivery. Additionally, we discuss the possibilities for replication in the area of population differences, in particular, age groups and more lasting effects as would be found in delayed post-tests. Finally, we point out another means of gathering information about thought processes and suggest the value of adding stimulated recalls to the existing data.

Algorithm for Computing Attribute Reduction Based on Radix Sort of Optimized Linked List Structure

The attribute reduction algorithm of radix sort to integer digit sequence table is not ideal. In this paper, based on how integers is stored in computer memory, low and high storage mode is designed for the solving algorithm of U/C chain structure of a new reasonable optimization. The time complexity of the algorithm from O (K|C| |U|) is reduced to O (K|C| |U| log (|M|)). The efficiency of the algorithm is improved by 3.8%.

Elements of mathematical phenomenology of self-organization nonlinear dynamical systems: Synergetics and fractional calculus approach

The modern dynamical systems of various physical natures, such as natural, social, economic, and technical ones, are complexes of various subsystems. They are connected by processes of intensive dynamic interaction and exchange of energy, matter, and information and incorporate nonlinear dynamics, memory, complicated transients, bifurcation and chaotic motion modes. Particularly, synergetics as a very young discipline deals with complex systems, i.e. it is concerned with the spontaneous formation of macroscopic spatial, temporal, or functional structures of systems via self-organization and is associated with systems composed of many subsystems, which may be of quite different natures. Synergetics take into account deterministic processes as treated in the dynamic systems theory including bifurcation theory, catastrophy theory, as well as basic notions of the chaos theory and develops its own approaches. Here, the fundamental basis of nonlinear theory of system׳s synthesis based on synergetics as well as fractional calculus approach in modern control theory together with its application will be presented. The difference of synergetic approach from the classical scientific methods is in identification of the fundamental role of self-organization in nonlinear dynamic systems and it is necessary to keep the conceptual correspondence to the main qualities of self-organization: nonlinearity—open systems—coherence. Synergetic approach is based on the natural homeostatic-conservation of the internal qualities of the dynamic systems of various natures. Namely, Russian scientist A.A. Kolesnikov developed a novel synergetic approach based on the ideas of modern mathematics, cybernetics, and synergetics to the synthesis of control systems for nonlinear, multidimensional and multilinked dynamic systems of various natures. The synergetic approach to control theory (synergetic control theory-SCT) is a novel nonlinear control method where the nonlinearities of a system are considered in the control design and a systematic design procedures. The invariants (synergies) and attractors, introduced as the main element of SCT, allow establishing direct link to the energy conservation laws, i.e. to the fundamental qualities of various objects. So, invariants, self-organization, and cascade synthesis are the fundamental notions of the SCT determining its essence, novelty, and content. Also, fractional calculus (FC) has a long history of three hundred years, over which a firm theoretical foundation has been established. All fractional operators consider the entire history of the process being considered, thus being able to model the non-local and distributed effects often encountered in natural and technical phenomena and they provide an excellent instrument for description of the memory, heredity, non-locality, self-similarity, and stochasticity of various materials and processes. Fractional dynamics can be encountered in various nonlinear dynamical systems such as visco-elastic materials, electrochemical processes, thermal systems, transmission and acoustics, chaos and fractals, biomechanical systems, and many others. The fractional dynamic systems with nonlinear control represent a relatively new class of applications of the FC which certified the FC as being a fundamental tool in describing the dynamics of complex systems as well as in advanced nonlinear control theory.

Demonstration of Unified Memory in FinFETs

Floating-body-induced transient mechanism in advanced FinFETs was investigated for unified and multi-bit memory capability. Nonvolatile memory operation was achieved by modifying the SOI buried insulator (BOX) such as the SiO 2- Si 3 N 4- SiO 2 (ONO) BOX can accumulate permanent charges. Charges are injected/removed in the Si 3 N 4 layer by back-gate or drain bias and sensed remotely, by gate coupling, through the modulation of the drain current flowing at the front interface. On the other hand, the isolated silicon body of the transistor can store volatile charges, generated by impact ionization and able to modulate the drain current flowing at the back interface. Our experimental results successfully demonstrate that these two different memory modes can be advantageously combined for multi-bit volatile memory operation. The volatile memory behavior strongly depends on the distribution of the nonvolatile charges stored in the nitride buried layer. Our measurements manifest that the nonvolatile charges located near the drain terminal have larger influence on the volatile memory operation than the charges located at the opposite terminal. Also, we reveal that the bias conditions and device geometry are important factors for the two memory modes.

RECENT TRENDS IN SPINTRONICS-BASED NANOMAGNETIC LOGIC

With the growing concerns of standby power in sub-100-nm CMOS technologies, alternative computing techniques and memory technologies are explored. Spin transfer torque magnetoresistive RAM (STT-MRAM) is one such nonvolatile memory relying on magnetic tunnel junctions (MTJs) to store information. It uses spin transfer torque to write information and magnetoresistance to read information. In 2012, Everspin Technologies, Inc. commercialized the first 64Mbit Spin Torque MRAM. On the computing end, nanomagnetic logic (NML) is a promising technique with zero leakage and high data retention. In 2000, Cowburn and Welland first demonstrated its potential in logic and information propagation through magnetostatic interaction in a chain of single domain circular nanomagnetic dots of Supermalloy ( Ni 80 Fe 14 Mo 5 X 1, X is other metals). In 2006, Imre et al. demonstrated wires and majority gates followed by coplanar cross wire systems demonstration in 2010 by Pulecio et al. Since 2004 researchers have also investigated the potential of MTJs in logic. More recently with dipolar coupling between MTJs demonstrated in 2012, logic-in-memory architecture with STT-MRAM have been investigated. The architecture borrows the computing concept from NML and read and write style from MRAM. The architecture can switch its operation between logic and memory modes with clock as classifier. Further through logic partitioning between MTJ and CMOS plane, a significant performance boost has been observed in basic computing blocks within the architecture. In this work, we have explored the developments in NML, in MTJs and more recent developments in hybrid MTJ/CMOS logic-in-memory architecture and its unique logic partitioning capability.

Real Time Accidents Diagnosis for Research Reactors using Adaptive Resonance Network

Real time locating faults in any nuclear research reactors plants are the highest importance requirements, aimed for safety of human and environmental reasons. Since a great fault can occur in a few milliseconds, accordingly, there is an increasing demand for automated systems to diagnose such failures. Adaptive Resonance network (ART) (1) is a neurofuzzy network, which is an important family of competitive neural learning model. Its memory mode is very similar to that of biological one, and memory capacitycan increase while the learning patterns increase. It can perform real-time online learning, and can work under non stationary world. In this research a new proposed a neural network classifier based on ART, which achieved preferable results than several other neural algorithms will be presented. The proposed algorithm obtains and diagnosis faults accidents patterns in the Multi- Purpose Research Reactor of Egypt, to avoid the risk of occurrence of a nuclear accident.

Selective switching mechanism in virtual machines via support vector machines and transfer learning

Virtualization is an essential technology in data centers allowing for a single machine to be used for multiple applications or users. With memory virtualization, two approaches, shadow paging (SP) and hardware-assisted paging (HAP), are taken by modern virtual machine memory managers. Neither memory mode is always preferred; previous studies have proposed to exploit the advantages of both modes by dynamically switching between these two paging modes based on the on-the-fly system behavior. However, the existing scheme makes the switching decision based on manual rules summarized for a specific architecture. This paper employs a machine learning approach that learns a decision model automatically and thus can adapt to different systems. Experimental results show that the performance of our switching mechanism can match or outperform either SP or HAP alone. Also, the results demonstrate that a machine learning-based decision model can match the performance of the hand-tuned model. Moreover, we further show that different hardware/software settings can affect on-the-fly system behavior and thus demand different decision models. Our scheme yields two effective decision models on two different machines. Additionally, transfer learning was used in order to efficiently train a model when faced with a new hardware configuration with only a limited number of training samples from the new machine.

Digital memory versatility of fully π-conjugated donor-acceptor hybrid polymers.

The fully π-conjugated donor-acceptor hybrid polymers Fl-TPA, Fl-TPA-TCNE, and Fl-TPA-TCNQ, which are composed of fluorene (Fl), triphenylamine (TPA), dimethylphenylamine, alkyne, alkyne-tetracyanoethylene (TCNE) adduct, and alkyne-7,7,8,8-tetracyanoquinodimethane (TCNQ) adduct, were synthesized. These polymers are completely amorphous in the solid film state and thermally stable up to 291-409 °C. Their molecular orbital levels and band gaps vary with their compositions. The TCNE and TCNQ units, despite their electron-acceptor characteristics, were found to enhance the π-conjugation lengths of Fl-TPA-TCNE and Fl-TPA-TCNQ (i.e., to produce red shifts in their absorption spectra and significant reductions in their band gaps). These changes are reflected in the electrical digital memory behavior of the polymers. Moreover, the TCNE and TCNQ units were found to diversify the digital memory modes and to widen the active polymer layer thickness window. In devices with aluminum top and bottom electrodes, the Fl-TPA polymer exhibits stable unipolar permanent memory behavior with high reliability. The Fl-TPA-TCNE and Fl-TPA-TCNQ devices exhibit stable unipolar permanent memory behavior as well as dynamic random access memory behavior with excellent reliability. These polymer devices were found to operate by either hole injection or hole injection along with electron injection, depending on the polymer composition. Overall, this study demonstrated that the incorporation of π-conjugated cyano moieties, which control both the π-conjugation length and electron-accepting power, is a sound approach for the design and synthesis of high-performance digital memory polymers. The TCNE and TCNQ polymers synthesized in this study are highly suitable active materials for the low-cost mass production of high-performance, polarity-free, programmable, volatile, and permanent memory devices that can be operated with very low power consumption, high ON/OFF current ratios, and high reliability.

A specific role for hippocampal mossy fiber's zinc in rapid storage of emotional memories.

We investigated the specific role of zinc present in large amounts in the synaptic vesicles of mossy fibers and coreleased with glutamate in the CA3 region. In previous studies, we have shown that blockade of zinc after release has no effect on the consolidation of spatial learning, while zinc is required for the consolidation of contextual fear conditioning. Although both are hippocampo-dependent processes, fear conditioning to the context implies a strong emotional burden. To verify the hypothesis that zinc could play a specific role in enabling sustainable memorization of a single event with a strong emotional component, we used a neuropharmacological approach combining a glutamate receptor antagonist with different zinc chelators. Results show that zinc is mandatory to allow the consolidation of one-shot memory, thus being the key element allowing the hippocampus submitted to a strong emotional charge to switch from the cognitive mode to a flashbulb memory mode. Individual differences in learning abilities have been known for a long time to be totally or partially compensated by distributed learning practice. Here we show that contextual fear conditioning impairments due to zinc blockade can be efficiently reduced by distributed learning practice.

Data Parallel and Scheduling Mechanism Based on Petri Nets

Data-parallel and task-parallel methods are the basic methods frequently used for algorithm design in parallel computing. Data-parallel method as name means is used for partition data to be processed into some small blocks considering storage and computing capacity such as memory size of a computation node, node number to take part in parallel computing and total data size, and etc. On the other hand, data dispensing strategy is an important problem carefully considered to increase the efficiency of computation. According to the characteristics of analysis of digital terrain, petri nets is introduced to describe the parallel relationships within data partitions based on data granularity model considering two kinds of computing modes, shared memory and distributed memory respectively, and corresponding scheduling algorithms are proposed for load balance. The experimental results show that our method is very usable to data partition and dispensation, in particular to distributed memory mode.

Bidirectional thermotaxis in Caenorhabditis elegans is mediated by distinct sensorimotor strategies driven by the AFD thermosensory neurons

The nematode Caenorhabditis elegans navigates toward a preferred temperature setpoint (Ts) determined by long-term temperature exposure. During thermotaxis, the worm migrates down temperature gradients at temperatures above Ts (negative thermotaxis) and performs isothermal tracking near Ts. Under some conditions, the worm migrates up temperature gradients below Ts (positive thermotaxis). Here, we analyze positive and negative thermotaxis toward Ts to study the role of specific neurons that have been proposed to be involved in thermotaxis using genetic ablation, behavioral tracking, and calcium imaging. We find differences in the strategies for positive and negative thermotaxis. Negative thermotaxis is achieved through biasing the frequency of reorientation maneuvers (turns and reversal turns) and biasing the direction of reorientation maneuvers toward colder temperatures. Positive thermotaxis, in contrast, biases only the direction of reorientation maneuvers toward warmer temperatures. We find that the AFD thermosensory neuron drives both positive and negative thermotaxis. The AIY interneuron, which is postsynaptic to AFD, may mediate the switch from negative to positive thermotaxis below Ts. We propose that multiple thermotactic behaviors, each defined by a distinct set of sensorimotor transformations, emanate from the AFD thermosensory neurons. AFD learns and stores the memory of preferred temperatures, detects temperature gradients, and drives the appropriate thermotactic behavior in each temperature regime by the flexible use of downstream circuits.

DC and AC Characterization of MgO Magnetic Tunnel Junction Sensors

We have fabricated multiple MgO magnetic tunnel junctions (MTJ) with 1.7 nm oxide, which are connected in series, and layed out in a serpentine geometry. We performed DC tunnel magnetoresistance measurements and AC impedance spectroscopy with crossed DC magnetic fields in the easy and hard axis directions. A simple <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$RLC$</tex></formula> circuit model is used to fit our data and characterize the dependence on capacitance <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$(C)$</tex></formula> and inductance <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$(L)$</tex></formula> of magnetization orientation of the MTJ sensors. We have found in our samples that <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$C$</tex></formula> and <formula formulatype="inline" xmlns:mml="http://www.w3.org/1998/Math/MathML" xmlns:xlink="http://www.w3.org/1999/xlink"><tex Notation="TeX">$L$</tex></formula> are higher in antiparallel than in parallel configuration. We discuss possible reasons for the existence of this field dependence and show the evolution of magnetic field vs. capacitance curve from memory mode into sensing mode at high frequencies.

Working Memory and Default Mode Network abnormalities in unaffected siblings of schizophrenia patients

BackgroundImpaired working memory (WM) is a hallmark of schizophrenia. In addition to classical WM regions such as the dorsolateral prefrontal cortex (DLPFC) and the striatum, dysfunctions in the default-mode network (DMN) contribute to these WM deficits. Unaffected siblings of patients also show WM impairments. However, the nature of the functional deficits underlying these impairments is unclear, mainly because of impaired performance confounding neuroimaging results. MethodsHere, we investigated WM and DMN activity in 23 unaffected siblings of schizophrenia patients and 24 healthy volunteers using fMRI and a Sternberg WM task. WM load was determined prior to scanning to ensure 90% accuracy for all subjects. ResultsSiblings showed hyperactivation during the encoding phase of WM in the right medial prefrontal cortex (MPFC) which is the anterior part of the DMN. No differences were found during the maintenance phase. During the retrieval phase, siblings showed hyperactivation in WM regions: DLPFC, inferior parietal cortex and the striatum. Siblings who showed hyperactivity in the MPFC during encoding showed DLPFC and striatum hyperactivation during retrieval. ConclusionsOur finding of hyperactivation in WM and DMN areas indicates that siblings fail to adequately inhibit DMN activity during demanding cognitive tasks and subsequently hyperactivate WM areas. This failure may reflect dopamine hyperactivity in the striatum which prevents adequate DMN suppression needed for effective WM. This study provides support for the notion that aberrant WM and DMN activation patterns may represent candidate endophenotypes for schizophrenia.

Multi-Bit Unified Memory Concept in FinFETs with ONO Buried Insulator

Transient mechanisms in SOI FinFETs fabricated on ONO buried insulator were investigated for multi-bit unified memory operation. In nonvolatile memory applications, charges are trapped/detrapped in the nitride buried layer by back-gate or drain bias and are sensed at the front-channel. Volatile charges are generated by impact ionization and stored in the silicon body of the transistor. The volatile charges are sensed at the back-channel. Our experimental results demonstrate multi-bit unified memory by advantageously combining the two memory modes. Systematic measurements reveal that the impact of the nonvolatile charges located near the drain side on the volatile sensing margin is larger than the charges located at the opposite side.

Recalling positive self-defining memories in depression: The impact of processing mode

Recalling positive memories is a powerful and effective way to improve mood. However, unlike never-depressed individuals, those with current or past depression do not benefit emotionally from positive memory recall. To examine whether rumination is involved in this difficulty, 80 participants (26 currently depressed, 29 recovered depressed, and 25 never depressed) were instructed to recall a positive self-defining memory while in a sad mood. They were then instructed to think about their memory, adopting either an abstract or concrete processing mode. Never-depressed and recovered depressed participants experienced improved mood after memory recall, regardless of processing mode. However, for depressed individuals neither an abstract nor a concrete processing mode produced emotional benefit. These findings suggest that a complex relationship exists among processing mode, memory type, and depressive status, and indicate that the way in which individuals process positive emotional material may have important consequences for treatment.

Cholesteric liquid crystal–carbon nanotube composites with photo-settable reversible and memory electro-optic modes

The photoresponsive electro-optical composites based on cholesteric liquid crystal (CLC) with optically controlled chirality and a minute amount of carbon nanotubes (CNTs) are studied. In cells with homeotropic anchoring, these composites exhibit a transition from fingerprint texture to homeotropic nematic texture in the course of photoinduced unwinding of the cholesteric helix. Compared with the CLC counterpart, this transition is much delayed, because of the stabilization of cholesteric filamentary domains by CNTs. The CLC-CNT composites demonstrate dual-mode operation with optical switching between reversible and memory mode. It is found that the memory response is associated with the elastic network of filamentary cholesteric domains that stabilizes the planar CLC texture reached in an electric field. In turn, the reversible mode corresponds to the unwound cholesteric state. Potential applications of this effect are discussed.

P.86: A Novel Bistable‐LCD Having Memory Display Mode and High‐Speed Switching Mode

AbstractA novel concept LCD which can be operated by two functions in one LCD has been proposed. One of the operating modes is a permanent memory mode without any applied voltage; images are rewritten by applying the vertical and horizontal electric field. The other one is a high‐speed switching mode which is operated by applying the horizontal electric field.

FinFETs with ONO BOX for multi-bit unified memory

Advanced FinFETs are fabricated on the ONO buried insulator.The nitride buried insulator can trap charges for the flash memory operation.The same device was investigated for the 1T-DRAM operation.By combining nonvolatile and volatile, multi-bit 1T-DRAM operation was obtained.Analog/logic and memory operation can be performed at the same cell. Floating-body-induced transient phenomena in advanced FinFETs fabricated on SOI with SiO2-Si3N4-SiO2 (ONO) buried insulator were studied for unified memory with multi-bit capability. The nitride layer can store nonvolatile charges by applying back-gate or drain bias. The trapped/detrapped charges in the nitride layer are sensed remotely, by gate coupling, through the variation of the drain current flowing at the front-channel. On the other hand, the transistor silicon body can accommodate volatile charges (generated by impact ionization) that also modify the drain current. Our experiments demonstrate that these two memory modes can be advantageously combined for multi-bit unified memory operation.