Search results for "activation entropy"

showing 2 items of 2 documents

Kinetik und Radikalausbeute beim Zerfall tertiärer Hyponitrite

1971

Der thermische Zerfall von Dicumylhyponitrit (1a) ist eine Reaktion erster Ordnung mit einer Aktivierungsenthalpie von 27.3 keal/Mol und einer Aktivierungsentropie von 8.5 cal/Mol·Grad in Isooctan. Die Radikalausbeute fur 1a betragt 84%, fur Di-tert.-butylhyponitrit 92%. Die Zerfallskonstanten substituierter Dicumylhyponitrite gehorchen bei Verwendung der σ+-Werte der Hammett-Beziehung mit σ=+ +0.34. Kinetic and Radical Efficiency for the Decomposition of Tertiary Hyponitrites The thermal decomposition of dicumylhyponitrite (1a) is a first order reaction with an activation enthalpy of 27.3 kcal/mole and an activation entropy of 8.5 e. u. in isooctane. The efficiency of radical formation is …

Inorganic ChemistryReaction rate constantHammett equationChemistryThermal decompositionFirst-order reactionEnthalpyPhysical chemistryActivation entropyKinetic energyDecompositionChemische Berichte
researchProduct

Studying the evolution of neural activation patterns during training of feed-forward ReLU networks

2021

The ability of deep neural networks to form powerful emergent representations of complex statistical patterns in data is as remarkable as imperfectly understood. For deep ReLU networks, these are encoded in the mixed discrete–continuous structure of linear weight matrices and non-linear binary activations. Our article develops a new technique for instrumenting such networks to efficiently record activation statistics, such as information content (entropy) and similarity of patterns, in real-world training runs. We then study the evolution of activation patterns during training for networks of different architecture using different training and initialization strategies. As a result, we see …

MultidisciplinaryArtificial IntelligenceElectronic computers. Computer sciencefeed-forward networksQA75.5-76.95activation patterns004 Informatikneural activationsRELUactivation entropy004 Data processingOriginal Research
researchProduct