Lens provides a number of example networks to give you a convenient reference and some fun toy problems to play with.
If this is your first time using Lens, you will want to explore the...
Feed-Forward Networks | |
xor.in | What simulator would be complete without the old XOR problem? |
digits.in | A very simple digit recognition task. |
hand-digits.in | A real handwritten digit recognition problem. |
bench.in | Used for benchmarking the simulator. Not very interesting. |
encoder.in | Another classic: the encoder network. |
negation.in | The bit negation problem from Vol. 1 of the PDP books. |
pick-one.in | Can the network find the culprit in the crowd? |
t-c.in | The T-C recognition problem from Vol. 1 of the PDP books. |
task.in | A gating network that learns to solve a specified task. |
logic.in | A network that solves boolean formulae with three variables. |
Simple Recurrent Networks | |
filler.in | A memory task. |
filler-srbptt.in | The same task using an SRBPTT network. |
adder2.in | A network that learns to add binary numbers. |
adder10.in | A network that learns to add base-10 numbers. |
sequence.in | A sequence prediction problem. |
rhw-sequence.in | The sequence task from Vol. 1 of the PDP books. |
logic2.in | The network must solve complex, nested boolean formulae. |
recoder.in | Can the network invent a language? |
Fully Recurrent Networks | |
integrate.in | Study the behavior of different types of integrating units. |
rand10x40.in | A feed-forward random mapping task with a continuous network. |
on-off.in | A continuous network that must turn units on and then off. |
Other Networks | |
jets-n-sharks.in | The Jets and Sharks interactive-activation network. |
kohonen.in | A Kohonen self-organizing map (SOM). |
boltz.in | Two simple Boltzmann machines on the digits task. |
boltz-complete.in | A pattern-completion task with Boltzmann machines. |