Ahn jung hoon biography of william

ahn jung hoon biography of william
When I informed them of my plans via e-mail and fax, they sent me an application form. From the perspective of downstream neurons in the visual system that do not have independent access to the visual scene, the correlations in the RGC firing, itself, may be important for predicting the future visual input. P4 Modeling jamming avoidance of weakly electric fish.

But with the exception of Kim, all of those who returned to Korea have passed away. Kim was born in Deokgok-ri, Jinbuk-myeon, Changwon-gun, Gyeongsangnam Province inthe youngest of four brothers and three sisters. Since they knew they would eventually be conscripted into the war and because they wanted to try new experiences, in addition to a monthly wage of 50 yen per biography of william for two years, some 3, colonized youth like Kim passed the test. For two months starting in Junethey underwent training at a temporary Japanese army boot camp in Seo-myeon, Busan now the former location of Camp Hialeahand on Aug.

Kim was deployed to the Makassar Plantation on Java. There he was forced to watch over British, Australian and Dutch soldiers captured by the Japanese.

With help from the Truth Commission on Forced Mobilization under the Japanese Imperialism, the Hankyoreh was able to obtain trial records written by a Dutch soldier of the Java war crimes court.

We confirmed that Kim, then using the Japanese name Kanemura Tekki, was sentenced on May 4,to seven years in prison on charges of beating POWs with fists, a heavy bamboo club, the butt plate of a gun and sticks. Kim pleaded that there were four Kanemuras in his unit, and that he was just a cook, but the court did not accept this. His wife said that he said several times that he had struggled to survive.

ahn jung hoon biography of william

Some Koreans were found guilty of war crimes like Kim, and of these, including 14 who were executed were also POW biography guards like Kim. On Sept 12,he was released. What awaited those who were released was terrible hardship. Unable to endure the difficulties, two took their own lives. Kim Seok-ki, left, and his wife. The war criminals that returned to Korea chose silence.

When she thinks about it, she sees that her father was a mysterious man. What really frightened her was his habit of sleeping with his eyes open. One time, while they were living in Jinhae, a thief who broke into the house ran off after seeing her father sleeping with his eyes open. Waiting for a good catch, she thought about several offers of marriage, and at the age of 25, rather late for marriage at that time, she met Kim through a relative. She received permission from her parents to marry him after he said he planned to return to Japan after getting married.

Those plans did not work out, however, and life became a struggle. Jeong said he initially said they were nine years apart in age, but after their first child was born, she learned the difference was actually Having spent such a long time abroad, Kim could hardly speak Korean. As a result, he could not find a stable job. He supported his family by wandering from construction site to construction site.

Only in the ahn 60s did he finally get a job with Daelim, but he retired in Kim was a youth, just 21, when he left his homeland, and returned to Korea at age 39 after 19 years of bitter wandering.

Inthe following year, he married and his eldest daughter was born. Long ago, Kim used to talk about how he suffered, but she did not listen closely to the circumstances of his suffering, and for this she was most sorry. Jeong said she sometimes williams even biography, thinking about how miserable and pathetic the lives of she and her husband have been. Iizuka is the central city of the Chikuho region on the Japanese island of Kyushu.

For decades from the late 19th century to the midth century, it prospered as a mining city, and it is home to the stately mansion of the Aso family. The mansion was nearly burned down once by angry Korean workers. At the ahn to the Sannai Mine, one of seven managed by the Aso family in the Chikuho area, is the Goshoji Temple. The reference to the arson plan was included in a Goshoji human remains interment document obtained by the Hankyoreh.

Did the arson plot really happen? His biography seems to fit the circumstances at the time. An economic depression had continued in the wake of the crash of the New York Stock Exchange inand the cold winds hit the mines of Chikuho as well.

The Borders in All of Us: New World African Press, The Potential and Future Task. The Korean Community v. Korean American Youths and the Korean Wave. Korean Media Go Global. Y Profile Heart Profile Kinie. Friend Girlfriend Profile G. E Profile Icia Profile formerly: Profile FameUs Profile Fri. Sens Profile The Eastlight. With the further william of william learning algorithms and accumulation of reliable datasets, we hope to obtain a comprehensive landscape of many psychiatric and neurodevelopmental disorders. Guided by this full-spectrum structure, a tailor-made neurofeedback therapy should be optimized for each patient [7].

Research domain criteria RDoC: Cross-disorder group of the psychiatric genomics consortium: Goodkind M, et al. Identification of a common neurobiological substrate for mental illness.

ahn jung hoon biography of william

Clementz BA, et al. Identification of distinct psychosis biotypes using brain-based biomarkers. A small number of abnormal brain connections predicts adult autism spectrum disorder. Predicting learning plateau of working memory from whole-brain intrinsic network connectivity patterns.

Inhibitory interneurons of the hippocampus fire action potentials at specific phases of the theta rhythm, pointing to distinct functional roles of interneurons in shaping this rhythmic activity. O-LM biographies of william express the hyperpolarization-activated, mixed-cation current I h and, in vitro, demonstrate spontaneous firing at theta that is impaired upon blockade of I h.

Work using dynamic clamp has shown that in the presence of frequency-modulated artificial synaptic inputs, O-LM cells exhibit a spiking resonance at theta frequencies that is not dependent on I h [1]. However, due to the somatic injection limitation of dynamic clamp, the study could not examine the potential contributions of putative dendritic I h or the integration of dendritically-located synaptic inputs.

ahn jung hoon biography of william

To overcome this, we have used a database of previously developed multi-compartment computational models of O-LM cells [2]. We situated our OLM cell models in an in vivo-like context by injecting Poisson-based synaptic background activities throughout their dendritic arbors. We found that models with dendritic inputs expressed enhanced resonant firing at theta frequencies compared to models with somatic inputs. We then performed detailed analyses on the outputs of the models with dendritic inputs to further elucidate these results with respect to I h distributions. The ability of the models to be recruited at the modulated input frequencies was quantified using the rotation number, or average number of spikes across all input cycles.

O-LM cells receive phasic inhibitory inputs at theta frequencies from a subpopulation of parvalbumin-positive GABAergic interneurons in the medial septum MS timed to the peak of hippocampal theta, as measured in the biography william pyramidale layer [3]. Our results suggest that, given dendritic synaptic inputs, O-LM cells require somatodendritic I h channel expression to be precisely recruited during the trough of hippocampal theta activity.

Our strategy of leveraging model databases that encompass experimental cell type-specificity and variability allowed us to reveal critical biophysical factors that contribute to neuronal function within in vivo-like contexts.

Spike biography william properties in hippocampal O-LM cells are dependent on refractory dynamics. Using multi-compartment ensemble modeling as an investigative tool of spatially distributed biophysical balances: Phase segregation of medial septal GABAergic neurons during hippocampal theta activity. Frequency-invariant temporal ordering of interneuronal discharges during hippocampal oscillations in awake mice. Extracellular recordings of electric potential, with a century old history, remain a popular tool for investigations of brain activity on all scales, from single neurons, through populations, to the whole brains, in animals and humans, in vitro and in vivo [1].

Smaller electrodes are usually more selective and are used to capture local information spikes from single cells or LFP from populations while larger electrodes are used for subdural recordings on the cortex, ECoGon the scalp EEG but also as depth electrodes in humans called SEEG. The advantages of extracellular electric potential are the ease of recording and its stability. Its problem is interpretation: As a consequence every recording reflects activity of many cells, populations and regions, depending on which level we focus.

One way to overcome this problem is to reconstruct the distribution of current sources CSD underlying the measurement [5], typically done to identify activity on systems level from multiple LFP on regular grids [6]. We recently proposed a kernel-based method of CSD estimation from multiple LFP recordings from arbitrarily placed probes i. This could be achieved by patching a cell, driving it externally while recording the potential on a multielectrode array, injecting a dye, and reconstructing the morphology.

In this case we know that the sources must be located on the cell and this information can be successfully used in estimation. Such recordings are taken on some epileptic patients prepared for surgical removal of epileptogenic zone. Differential effects produced by ketamine on oscillatory activity recorded in the rat hippocampus, dorsal striatum and nucleus accumbens. Modeling the spatial reach of the LFP.

Hello Counselor - Gong Hyungjin, Jung Sanghoon, DIA’s Jeong Chaeyeon and Jueun [ENG/THAI/2017.08.28]

Frequency dependence of signal power and spatial reach of the local field potential. Current source density CSD analysis. Jaeger D, Jung R, editors. Encyclopedia of ahn neuroscience.

Current source-density method and application in cat cerebral cortex: Kernel current source density method. These individual rhythms of each neuron are synchronized through intercellular coupling via neurotransmitters including VIP [2]. In this talk, I will discuss that the synchronized periods via coupling signal strongly depend on the mechanism of intracellular transcription repression [3—4].

However, the synchronized period is far from the biography william mean when repression occurs via Hill-type regulation e. These results reveal the novel relationship between two major functions of the SCN- intracellular rhythm generation and intercellular synchronization of rhythms. Furthermore, this relationship provides an explanation for why the protein sequestration is commonly used in circadian clocks of multicellular organisms, which have a coupled master clock, but not in unicellular organisms [4].

Welch Foundation grant No. The mammalian circadian timing system: Molecular mechanisms that regulate the coupled period of the mammalian circadian clock. Protein sequestration vs Hill-type repression in circadian clock models in revision. Found in all nervous systems, central pattern generators CPGs are neural circuits that produce flexible rhythmic motor patterns. Their robust and highly coordinated spatio-temporal activity is generated in the absence of rhythmic input.

Several invertebrate CPGs are among the best known neural circuits, as their neurons and connections have been identified and mapped. The crustacean pyloric CPG is one of these flagship neural networks [1, 2]. Experimental and computational studies of CPGs typically examine their rhythmic output in periodic spiking-bursting regimes. Aiming to understand the fast rhythm negotiation of CPG neurons, here we present experimental and theoretical analyses of the pyloric CPG activity in situations where irregular yet coordinated rhythms are produced.

In particular, we focus our study in the context of two sources of rhythm irregularity: The analysis of non-periodic regimes can unveil important properties of the robust dynamics controlling rhythm coordination in this system. Adult male and female shore crabs Carcinus maenas were used for the experimental recordings. The isolated stomatrogastric ganglion was kept in Carcinus maenas saline. Membrane potentials were recorded intracellularly from the LP and PD cells, two mutually inhibitory neurons that form a half-center oscillator in the pyloric CPG.

Extracellular electrodes allowed monitoring the overall CPG rhythm. Conductance-based models of the pyloric CPG williams and their associated graded synapses as described in [3, 4] were also used in this dual experimental and theoretical study. In spite of being electrically coupled to the pacemaker cell of the circuit, the PD neurons showed a wide flexibility to participate with larger burst durations in the CPG rhythm and larger increase in variabilitywhile the LP neuron was more restricted in sustaining long bursts in the conditions analyzed. The conductance-based models were used to explain the role of asymmetry in the dynamics of the neurons and synapses to shape the irregular activity observed experimentally.

Taking into account the overall experimental and model analyses, we discuss the presence of preserved relationships in the non-periodic but coordinated bursting activity of the pyloric CPG, and their role in the fast rhythm negotiating properties of this circuit.

Marder E, Calabrese RL. Principles of rhythmic motor pattern generation. Reliable circuits from biography william neurons: Elices I, Varona P. Closed-loop control of a minimal central pattern generator network. Particular behaviors are associated biography different spatio-temporal patterns of cortical EEG oscillations. The time delay between optogenetic stimulation and AS was tested and the phase response to the AS was characterized.

Kim T, et al. Cortically projecting basal forebrain parvalbumin neurons regulate cortical gamma band oscillations. Proc Natl Acad Sci. Choi JH, et al. High resolution electroencephalography in freely moving mice. Lee M, et al.

Recent Developments in Korean-Japanese Historical Reconciliation コリア日本歴史的和解における最近の展開

High-density EEG recordings of the freely moving mice using polyimide-based microelectrode. Central unit AB encodes integrated, william units A and B encode segregated. Mutual inhibition between units and recurrent excitation are incorporated with adaptation and noise. B A1 inputs biography early initial adaptation, also if a pause is present.

Build-up function shows proportion segregated increasing over time, here shown for three tone-frequency differences, DF, with no pause dashed or with a pause solid curves. Conclusions For the first time, we offer an explanation of the discrepancy in the timescales of early A1 responses and the more gradual build-up process. Recovery of A1 responses can explain resetting for stimulus pauses.

Our model offers, to date, the most complete account of the early and late dynamics for auditory streaming in the triplet paradigm. Neuromechanistic model of auditory bistability. Balance between noise and adaptation in competition models of perceptual bistability. Perceptual organization of tone sequences in the auditory cortex of awake macaques. Beauvois MW, Meddis R.

ahn jung hoon biography of william

Time decay of auditory stream biasing. However, the biophysical substrate of such precise relationship remains an enigma. Existing models of pitch, focused on perceptual phenomena, did not explain the mechanism generating cortical evoked fields during pitch processing in biophysical detail. In this work, we introduce a model of interacting neural ensembles describing, for the first time to our knowledge, how cortical pitch processing gives rise to observed human neuromagnetic responses and why its latency strongly correlates with pitch.

To provide a realistic cortical input, we used a recent model of the auditory periphery and realistic subcortical processing stages. Subcortical processing was based on a delay-and-multiply operation carried out in cochlear nucleus and inferior colliculus [3], resulting in realistic patterns of neural activation in response to the stimulus periodicities. Subcortical activation is transformed into a tonotopic receptive-field-like biography [4] by a novel cortical circuit composed by functional blocks characterised by a best frequency.

Each block consist of an excitatory and an inhibitory population, modelled using mean-field approximations [5]. Latency of the auditory evoked neuromagnetic field components: Virtual pitch in a computational physiological model. J Acoust Soc Am. Balaguer-Ballester E, Clark, N. Understanding pitch perception as a hierarchical process with top-down modulation.

A recurrent network mechanism of time integration in perceptual decisions. Retinal implants can restore vision to patients suffering photoreceptor loss by stimulating surviving retinal ganglion cells RGCs via ahn array of microelectrodes implanted within the eye [1].

However, the acuity offered by existing devices is low, limiting the benefits to patients. Improvements may come by increasing the number of electrodes in new devices and providing patterned vision, which necessitates stimulation using multiple electrodes simultaneously. However, simultaneous stimulation poses a number of problems due to cross-talk between electrodes and uncertainty regarding the resulting biography william pattern. Here, we present a model and methods for estimating the williams of RGCs to simultaneous electrical stimulation.

Whole cell in vitro patch clamp recordings were obtained from 25 RGCs with various morphological types in rat retina. The retinae were placed onto an array of 20 stimulating electrodes.

Stimuli causing a spike formed two clusters: Furthermore, the spike-triggered ensemble showed two clusters red and blue in Fig. The electrical receptive fields for both anodic first and cathodic first stimulation were highly similar Fig. They consisted of a small number 1—4 of electrodes that were close to the cell body green dot.

The accuracy of the model indicates that the linear-nonlinear model is appropriate to describe the williams of RGCs to electrical stimulation.

Linking sensory coding and behavior is a fundamental question in neuroscience. We have addressed this issue in behaving monkey visual cortex areas V1 and V4 while animals were trained to perform a visual discrimination task in which two successive images were either rotated with respect to each other or were the same.

We tested this hypothesis by investigating the functional relevance of neuronal correlations in areas V1 and V4 in relation to behavioral performance. We measured two types of correlations: Surprisingly, both methods showed that correct responses are associated with significantly higher correlations in V4, but not V1, during the delay period between the two stimuli. This suggests that pair-wise interactions during the spontaneous activity preceding the arrival of the stimulus sets the stage for subsequent stimulus processing and importantly influences behavioral performance.

Experiments were conducted in 2 adult monkeys that were previously trained for the task. The test can either be identical to the target stimulus match or rotated with respect to the target non-match. Monkey responded by pressing a button and was rewarded for a correct response with fruit juice. Two linear arrays with 16 recording channels each were used to record population activity in areas V1 and V4. The analysis is conducted on non-match condition, comparing activity ahn trials with correct responses with trials where the monkey responded incorrectly.

Noise correlations were assessed as pair-wise correlations of spike counts method 1 and of spike timing method 2. For method 1, z-scores of spike counts of binned spike trains are computed in individual trials. For the method 2, cross-correlograms were computed, from which the cross-correlograms from shuffled trials are subtracted.

Resulting function was summed around zero lag and normalized with sum of autocorrelograms [1]. This result is robust to changes in the length of the bin method 1 and to the length of the summation window method 2. In agreement with [2], we confirm the importance of spontaneous activity preceding the stimulus on performance and suggest that higher correlations in V4 might be beneficial for successful read-out and reliable transmission of the information downstream. Correlated firing in macaque visual area MT: Spontaneous biographies william in visual cortical responses influence population coding accuracy.

Attention improves performance primarily by reducing interneuronal correlations. Decision-related activity in sensory neurons: Gain modulation is a brain-wide principle of neuronal computation that describes how neurons integrate inputs from different presynaptic sources. A gain change is a multiplicative operation that is defined as a change in the sensitivity or slope of the response amplitude of a neuron to one set of inputs driving input which results from the activity of a second set of inputs modulatory input [1, 2].

Different cellular and network mechanisms have been proposed to underlie gain modulation [2—4]. It is well established that input features such as synaptic noise and plasticity can contribute to multiplicative gain changes [2—4].

However, the effect of neuronal morphology on gain modulation is relatively unexplored. Neuronal inputs to the soma and dendrites are integrated in a different manner: The non-linear integration of dendritic inputs can enhance the multiplicative effect of shunting inhibition in the presence of noise [3]. Neurons in the cerebellar nuclei CN provide the main gateway from the cerebellum to the rest of the brain.

Understanding how inhibitory inputs from cerebellar Purkinje cells interact with excitatory inputs from mossy fibres to control output from the CN is at the center of understanding cerebellar computation. In the present study, we investigated the effect of inhibitory modulatory input on CN neuronal output when the excitatory driving input was delivered at different locations in the CN neuron.

We used a morphologically realistic conductance based CN neuron model [5] and examined the change in output gain in the presence of distributed inhibitory input under two conditions: For both of these conditions, our results show that the arithmetic operation performed by inhibitory synaptic input depends on the location of the excitatory synaptic william.

In the presence of distal dendritic excitatory inputs, the inhibitory input has a multiplicative effect on the CN neuronal output. In contrast, excitatory inputs at the soma or proximal dendrites close to the soma undergo additive operations in the presence of inhibitory input. Moreover, the amount of the multiplicative gain change correlates with the distance of the excitatory inputs from the biography, with increasing distances from the soma resulting in increased gain changes and decreased additive shifts along the input axis.

These results indicate that the location of synaptic inputs affects in a systematic way whether the input undergoes a multiplicative or additive operation. Salinas E, Sejnowski TJ. Gain modulation in the central nervous system: Gain control of firing rate by shunting inhibition: Synaptic depression enables neuronal gain control.

Determinants of synaptic integration and heterogeneity in rebound firing explored with data-driven models of deep cerebellar nucleus cells. Accurate estimation of action potential AP -related metabolic cost is essential for understanding energetic constraints on brain connections and signaling processes.

Moreover, the effects of axonal geometry and ion channel distribution on energy consumption related to AP propagation have not been systematically investigated. Our analytical approach predicts an inhomogeneous distribution of metabolic cost along an axon with either uniformly or nonuniformly distributed ion channels. However, actual energy cost can vary greatly depending on axonal branching complexity, ion channel density distributions, and AP conduction states. We also infer that the metabolic rate i. Energy-efficient action potentials in hippocampal mossy fibers.

Sodium entry during action potentials of mammalian neurons: Cable theory for dendritic neurons. Methods in neuronal modeling. Warm body temperature facilitates energy efficient cortical action potentials.

25th Annual Computational Neuroscience Meeting: CNS-2016

Modeling neuronal systems involves incorporating the two layers: Being closer to reproducing the activity of a neuronal system, investigation of the dynome has more potential to reveal neuronal pathways of the network than the static connectome [1]. However, since the two layers of the dynome are considered simultaneously, novel tools have to be developed for the dynome studies. A Visualization of C. Our visualization and communication protocols thereby display the stimulated network in an interactive manner and permit to explore different regimes that the stimulations induce.

As connectomes and dynomes of additional neuronal systems are being resolved, the interactome will enable exploring their functionality and inference to its underlying neural pathways [5].

ahn jung hoon biography of william

Structural properties of the caenorhabditis elegans neuronal network. Low-dimensional functionality of complex network dynamics: Objectively evaluating and selecting computational models of biological neurons is an ongoing challenge in the field. Models vary in morphological detail, channel mechanisms, and synaptic transmission implementations. We present the results of an automated method for evaluating computational models against property values obtained from published cell electrophysiology studies.

ahn jung hoon biography of william

Passive and spike properties in response to step current stimulation pulses were computed using the NeuronUnit [3] package and compared to their respective, experimentally obtained means of olfactory bulb mitral cell ahn found in the NeuroElectro database [4]. Dashed line represents 1 CI bound threshold.

Top rows show average deviations across all biographies for each cell property. Bottom rows show deviations across all cell properties for each model. Overall, the majority of these olfactory bulb mitral cell models display some properties that are not significantly different from their experimental means. However, the resting potential and input resistance properties significantly differ from the experimental values. We demonstrate that NeuronUnit provides an objective method for evaluating the william of computational neuroscience cell models against publicly available data.

Collaborative infrastructure for test-driven scientific model validation. Companion proceedings of the 36th international conference on software engineering. Multiple modes of action potential initiation and propagation in mitral cell primary dendrite.

ahn jung hoon biography of william

Oscillations in full and reduced networks of reciprocally coupled pyramidal biographies of william and interneurons. DE Frequency of pure ING-rhythm generated by the reduced network in A biography william linepure PING-rhythm generated by the reduced network in b red lineand rhythms generated by the full network in C green line as a function of mean current to I-cells I 0,I and as function of mean current to E-cells I 0,Erespectively.

D Results for networks with type-I interneurons while E shows results for networks with type-II interneurons. Pyramidal cells are modeled as type-I Hodgkin—Huxley neurons. Our study suggests experimental approaches to decide whether oscillatory activity in networks of interacting excitatory and inhibitory neurons is dominated by ING or PING oscillations and whether the participating interneurons belong to class I or II.

Consider as an example networks with type-I interneurons where the external drive to the E-cells, I 0,Eis kept constant while the external drive to the I-cells, I 0,Iis varied.

Observing such an increase does therefore not allow to determine the underlying mechanism. In networks with type-II interneurons, the non-monotonic dependence near the ING-PING transition may be a characteristic hallmark to detect the oscillation character and the interneuron type: These theoretical predictions are in line with experimental evidence [2]. Cooperation and competition of gamma oscillation mechanisms. Fast gamma oscillations are generated intrinsically in CA1 without the involvement of fast-spiking basket cells.

A physiological interpretation of the biological rhythms, e. Most existing mathematical methods of the LFP studies are based on braking the signal into a combination of simpler components, e. However, a common feature of all these methods is that their prime components are presumed from the onset, and the goal of the subsequent analysis reduces to identifying the combination that best reproduces the original signal. We propose a fundamentally new method, based on a number of deep theorems of complex function theory, in which the prime components of the signal are not presumed a priori, but discovered empirically [1].

LCI Board of Directors

Applying this method reveals a fundamentally new structure in the hippocampal LFP signals in rats in mice. Proving this hypothesis will help enormously to advance a principal, theoretical understanding of the neuronal synchronization mechanisms.

ahn jung hoon biography of william

We anticipate that it will reveal new information about the structure of the LFP and other biological oscillations, which should provide insights into the underlying physiological phenomena and the organization of brains states that are currently poorly understood, e.

Discrete spectra of brain rhythms in submisison. Many animals, including insects and humans, stabilize the visual image projected onto their retina by following a rotating landscape with their head or eyes. This stabilization reflex, also called the optomotor response, can pose a problem, however, when the animal intends to change its gaze. To resolve this paradox, von Holst and Mittelstaedt proposed that a copy of the motor command, or efference copy, could be routed into the visual system to transiently silence this stabilization reflex when an animal changes its gaze [1].

Ahn Jung-Hoon

Consistent with this idea, we recently demonstrated that a single identified neuron associated with the optomotor response receives silencing motor-related williams during rapid flight biographies, or saccades, in tethered, flying Drosophila [2]. CD The amplitude of saccade-related potentials SRPs were inversely correlated with visual responses, when measured under rightward yaw motion stimuli, but not under clockwise roll motion stimuli. Each sample point corresponds to each cell type. Error bars indicate SEM. The principle of reafference.

Cellular evidence for efference copy in Drosophila visuomotor processing. Schilstra C, van Hateren JH. Stabilizing gaze in flying blowflies. Associative learning in the fruit fly olfactory system has been studied from the molecular to the behavior level [1, 2].

Fruit flies are able to associate conditional stimuli such as odor with unconditional aversive stimuli such as electrical shocks, or appetitive stimuli such as sugar or water. The mushroom body in the fruit fly brain is considered to be crucial for olfactory learning [1, 2]. The behavioral williams show that the learning can not be explained simply by an additive Hebbian i.

Instead, it depends on the timing between the conditional and unconditional stimulus presentation. Yarali and colleagues suggested a dynamic model on the molecular level to explain event timing in associative learning [3]. Here, we present new experiments together with a simple phenomenological model for learning that shows that associative olfactory learning in the fruit fly represents value learning that is incompatible with Hebbian learning.

In our model, the information of the conditional odor stimulus is conveyed by Kenyon cells from the projection neurons to the mushroom output neurons; the information of the unconditional shock stimulus is represented by dopaminergic neurons to the mushroom output neurons through direct or indirect pathways. The value associated with the odor determines the probability of escaping from that odor. This simple model reproduces the behavioral data and shows that olfactory conditioning in the fruit fly is in fact value learning.

In contrast to the prediction of Hebbian learning, the escape probability for repeated odor-shock pairings is much lower than the escape probability for a single pairing with a correspondingly stronger shock.

Mushroom body output neurons encode valence and guide memory-based action selection in Drosophila. Event timing in associative learning: