ARTIFICIAL NEURAL NETWORKS IN AKKA
summary
Shows that Akka actors implement small, timed neural circuits: actors act as actors as neurons with input buffers, a refractory window and a max interaction gap, plus three neuron-types—signal-sum, delay gate and silencing neuron. A compact network of roughly a dozen actors interprets short/long bursts (gaps matter not amplitudes) and reliably recognizes SOS in Morse by using winner‑takes‑all silencing and timing rather than heavy numeric layers. Code is on GitHub and the example works.
AI-written summary · about
chapters (19)
- 0:00Introduction
- 0:38Outline
- 1:37Inspiration
- 2:39Signals
- 3:52Nervous system
- 4:40Perfection model
- 5:34Reactive neural networks
- 6:08An actor
- 7:04Aka features
- 8:28Unordered dependent
- 9:09Re refractory period
- 10:18Signal
- 13:00Neurons
- 13:57Example
- 15:17Delay Gate
- 16:15Path to build the network
- 17:01Pattern recognition
- 18:23What we have learned
- 20:07Recommendations
related talks
47mNetworks and Types — the Future of Akka
40mREST no more - Using Actors for the Internet of LEGO...
22mAdvanced Akka features
47mAkka - Multithreaded Programming Without the Pain
32mAdopting Actors... an epic tale of loss and learning
35mOne year with Akka 2.6
37mBeyond Fault Tolerance
38mFresh from the Oven: Akka gets Typed