1,720,960 research outputs found
Memory consolidation in memristive systems
No full textThis thesis investigates the challenge of memory consolidation and learning in artificial synapses. The adoption and evolution of artificial intelligence (AI) also by products a frequently overlooked exponentially increasing need for information processing and data storage. This issue is either met with the physical expansion of storage facilities or with the inevitable forgetting of old information in favour of new; both of which seriously hinder the performance of embedded AI systems. This work presents a novel approach in emulating the complex biochemical mechanisms which allow neuronal synapses to store multiple memories on top of the other and at different timescales, like a palimpsest, and which give rise to the incredible learning capacity of biological intelligence. This work mainly focused on exploiting the intrinsic time dependent volatility in emerging memristive nanotechnologies to showcase palimpsest consolidation. Memristive volatility was studied using a data-driven approach and device-agnostic characterisation and mathematical modelling methods were developed to uncover the main properties of the mechanism. It was found that volatility can exist bidirectionally in TiO2 memristors and that its time constants can be manipulated via the invasiveness and/or frequency of device stimulation. Importantly, within a given observation time window, volatility was shown to operate at two timescales; a fast decay of large magnitude followed by a saturating steady state and a small non-volatile residue. By operating memristive devices as binary synapses, spiking plasticity events were able to store long-term memories in the non-volatile residue, while expressing the opposing state in the short-term. Palimpsest consolidation was examined in simulated memory networks which were able to protect long-term memories while expressing up to hundreds of uncorrelated short-term memories. It was also found that these networks bear close resemblance to the visual working memory of mammalian brains. The same plasticity dynamics were finally extended towards the context of neuronal activity detection, where memristive sensors were able to ’learn’ during high spiking frequencies and ’forget’ during less active timeframes. The results presented in this thesis verify the candidacy of volatile memristors as natural facilitators of learning in AI. The ability to learn continuously without catastrophically forgetting old memories, can create new possibilities in the way AI can be used to undertake more generalised tasks. Moreover, the same artificial synapses have shown immense potential in neural interfacing. This can potentially reshape the ways AI is currently interpreted and lead to novel research which aims to integrate both biological and artificial intelligence
Bidirectional volatile signatures of metal-oxide memristors-Part II: modeling
No full textVolatility in metal-oxide resistive random access memory (RRAM) families has mostly been treated as an unwanted side-effect, although recently there are trends to interpret such behavior as an additional technological feature. To date, the field has seen early demonstrations of possible applications that harness volatility. Moreover, some work has been conducted to understand both the mechanisms responsible for this behavior. In the context of modeling RRAM volatility, we still lack a comprehensive model that could allow simulations in a larger scale. In an attempt to fill this gap, this work presents a modeling framework that can account for RRAM relaxation characteristics. Specifically, we show how volatility can be simulated to significant accuracy when the resistive state (RS) of a device as well as the stimulus protocol in use are well-defined. Importantly, our approach is solely data-driven and decoupled from previous physical modeling studies on volatility. Our results work for both stimulation polarities and are consistent for a number of TiOx devices in use. Moreover, the mathematical relations that unfold via modeling volatility provide further intuition on the effect that invasive protocols can have on this technology. This modeling solution enables more advanced studying of memristive technologies in one hand, as well as more intricate designs of larger systems that can account for transient RRAM changes over time
Thermal effects on initial volatile response and relaxation dynamics of resistive RAM devices
No full textResistive RAM (RRAM) or memristors are a class of electronic device whose resistance depends on voltage history. The changes in resistance can be divided into two categories, volatile and non-volatile. To date, the characteristics of non-volatile switching have been explored extensively with volatile switching behaviour still remaining more obscure. Here we investigate the temperature effects on TiOx based memristor volatility, and integrate these observations into a previously developed model for volatile switching. We show how device temperature affects the magnitude of the volatile resistive state in response to input stimulation, as well as the corresponding relaxation time constant. Importantly, these effects are polarity dependent. This work is part of an effort towards building a more comprehensive model of RRAM behaviour covering volatile and non-volatile phenomena as well as various environmental effects on them
Palimpsest memories stored in memristive synapses
No full textBiological synapses store multiple memories on top of each other in a palimpsest fashion and at different time scales. Palimpsest consolidation is facilitated by the interaction of hidden biochemical processes governing synaptic efficacy during varying lifetimes. This arrangement allows idle memories to be temporarily overwritten without being forgotten, while previously unseen memories are used in the short term. While embedded artificial intelligence can greatly benefit from this functionality, a practical demonstration in hardware is missing. Here, we show how the intrinsic properties of metal-oxide volatile memristors emulate the processes supporting biological palimpsest consolidation. Our memristive synapses exhibit an expanded doubled capacity and protect a consolidated memory while up to hundreds of uncorrelated short-term memories temporarily overwrite it, without requiring specialized instructions. We further demonstrate this technology in the context of visual working memory. This showcases how emerging memory technologies can efficiently expand the capabilities of artificial intelligence hardware toward more generalized learning memories
cgiotis/palimpsest_memories: Dataset and Code Material for "Palimpsest Memories Stored in Memristive Synapses" Manuscript."
No full textDataset and code material for the "Palimpsest Memories Stored in Memristive Synapses" manuscript, currently available on ArXiv. This repository currently contains instructions and the necessary dataset to recreate Figures 1 and 2 from the paper, as well as parametric code to recreate the simulation results and generate the relevant plots (Figures 3 and 4).</span
Going Beyond Counting First Authors in Author Co-citation Analysis
Full text linkThe present study examines one of the fundamental aspects of author co-citation analysis (ACA) - the way co-citation
counts are defined. Co-citation counting provides the data on which all subsequent statistical analyses and mappings
are based, and we compare ACA results based on two different types of co-citation counting - the traditional type that
only counts the first one among a cited work's authors on the one hand and a non-traditional type that takes into
account the first 5 authors of a cited work on the other hand. Results indicate that the picture produced through this non-traditional author co-citation counting contains more coherent author groups and is therefore considerably clearer. However, this picture represents fewer specialties in the research field being studied than that produced through the traditional first-author co-citation counting when the same number of top-ranked authors is selected and analyzed. Reasons for these effects are discussed
Variations on the Author
Full text link“Variations on the Author” discusses two of Eduardo Coutinho’s recent films (Um Dia na Vida, from 2010, and Últimas Conversas, posthumously released in 2015) and their contribution to the general question of documentary authorship. The director’s filmography is characterized by a consistent yet self-effacing form of authorial self-inscription: Coutinho often features as an interviewer that rather than express opinions propels discourses; an interviewer that is good at listening. This mode of self-inscription characterizes him as an author who is not expressive but who is nonetheless markedly present on the screen. In Um Dia na Vida, however, Coutinho is completely absent form the image, while Últimas Conversas, on the contrary, includes a confessional prologue that moves the director from the margins to the center of his films. This article examines the ways in which these works stand out in the filmography of a director who offers new insights into the notion of cinematic authorship
Appropriate Similarity Measures for Author Cocitation Analysis
Full text linkWe provide a number of new insights into the methodological discussion about author cocitation analysis. We first argue that the use of the Pearson correlation for measuring the similarity between authors’ cocitation profiles is not very satisfactory. We then discuss what kind of similarity measures may be used as an alternative to the Pearson correlation. We consider three similarity measures in particular. One is the well-known cosine. The other two similarity measures have not been used before in the bibliometric literature. Finally, we show by means of an example that our findings have a high practical relevance.information science;Pearson correlation;cosine;similarity measure;author cocitation analysis
- …
