Journal: Computational Intelligence and Neuroscience

Article Title: Acceleration of Deep Neural Network Training Using Field Programmable Gate Arrays

doi: 10.1155/2022/8387364

Figure Lengend Snippet: The accuracy comparison for AlexNet model.

Article Snippet: It achieves 203.75 GOPS on Terasic DE1 SoC with the AlexNet model and 196.50 GOPS with the VGG-16 model on Terasic DE-SoC.

Techniques: Comparison

Journal: bioRxiv

Article Title: Creativity and memory: Cortical representational change along with amygdala activation predict the insight memory effect

doi: 10.1101/2023.06.13.544774

Figure Lengend Snippet: Identification of VOTC regions showing Representational Change (RC) Note. Values = estimated marginal means ± SEM; asterisk = statistical significance; dot represents trend without statistical significance. HI-I = solved trials with high insight. Panel A. RC from pre to post solution: Multivoxel pattern similarity. Multivoxel patterns per ROI for each time point (pre- and post solution) are extracted and subsequently correlated. Pre= 0.5sec after stimulus presentation; post = during solution button press. Those Pre-Post Solution Similarity values (r) are subsequently estimated in a linear mixed model as a function of insight and ROI. Bar plots show change (Δ) in Multivoxel Pattern Similarity (MVPS) analysis. Change in MVPS = 1 minus the correlation between the post and pre solution multivoxel pattern in the respective ROI. Panel B. RC from pre to post solution: Representational strength - AlexNet. This RSA method employs four steps. ( B1 ) A brain activation pattern matrix (APM, size 120×120) is generated for each region-of-interest (ROI) and each time point (pre- post solution, see Panel A) where each cell is representing a multivoxel pattern similarity value for each Mooney image pair. ( B2 ) A conceptual stimuli model (here using AlexNet, size 120×120) is generated where each cell is representing a similarity value for each Mooney object pair. ( B3 ) For each brain region and each time point, the row of each stimulus (∼120) in the stimuli model and in the APM are correlated yielding a representational strength measure (i.e. brain-model fit) per region and time point. ( B4 ) The representational strength is used as a dependent variable in linear mixed models to investigate which ROIs exhibit an insight-related increase in representational strength from pre- to post solution (time). Barplots show Representational Strength [Rep-Str] = “second order correlation” between multivoxel patterns in respective ROI at pre and post response time point and conceptual stimuli models (AlexNet and Word2Vec [W2V]).

Article Snippet: The linear trend of the insight-memory-factor significantly accounted for the correlation of activity pattern over both brain regions (pFusG, iLOC) with the conceptual AlexNet model during solution (Chi 2 (2)=9.66, p <.01, ß=.07), suggesting that the representational strength of the Mooney object during post solution is associated with insight-related better memory (see ).

Techniques: Activation Assay, Generated

Journal: bioRxiv

Article Title: Creativity and memory: Cortical representational change along with amygdala activation predict the insight memory effect

doi: 10.1101/2023.06.13.544774

Figure Lengend Snippet: Insight memory factor predicts multivariate activity in VOTC, univariate activity in amygdala and Amygdala-VOTC functional connectivity. Note . Values represent estimated marginal means ± (between subject) SEM. Forgotten = subsequently forgotten trials; Rem_LO-I: subsequently remembered trials originally solved with low insight; Rem_HI-I = subsequently remembered trials originally solved with high insight. Panel A: change (Δ) in Multivoxel Pattern Similarity (MVPS). Change in MVPS = 1 minus the correlation between the post and pre solution multivoxel pattern in the respective ROI. Panel B: Representational strength of solution object (post solution) is measured via a conceptual model created out of the penultimate layer of AlexNet. Panel C: Representational strength is measured via a conceptual Word2Vec [W2V] model. Panel D: Amygdala mean activity at post solution divided by insight memory conditions. Panel E: Functional connectivity between Amygdala and VOTC. Values represent averaged correlation coefficients between left and right amygdala and left and right iLOC or pFusG.

Article Snippet: The linear trend of the insight-memory-factor significantly accounted for the correlation of activity pattern over both brain regions (pFusG, iLOC) with the conceptual AlexNet model during solution (Chi 2 (2)=9.66, p <.01, ß=.07), suggesting that the representational strength of the Mooney object during post solution is associated with insight-related better memory (see ).

Techniques: Activity Assay, Functional Assay