Short interactive experiments
Energy, memory, probability, and hidden variables.
Each experiment isolates one central idea behind Hopfield networks, Boltzmann machines, and restricted Boltzmann machines. Use one during a lecture, or let people explore after scanning a QR code.
1. Boltzmann factor
Energy becomes probability
Change temperature and energy levels. See why low energy is favored and why the partition function is needed.
2. Hopfield network
Memory as an attractor
Corrupt a stored pattern, then let deterministic energy descent recover a nearby memory.
+1J-1
3. Boltzmann machine
From descent to sampling
Keep an Ising-like energy, add temperature, and watch stochastic Gibbs updates explore a distribution.
4. Restricted Boltzmann machine
Hidden causes create correlations
Tune a visible-hidden model for Anna, Lukas, and Bruno. Inspect probabilities and sample one Gibbs step.