Skip to content

Commit 5209176

Browse files
Documentation tweaks.
1 parent b04f778 commit 5209176

File tree

1 file changed

+5
-6
lines changed

1 file changed

+5
-6
lines changed

docs/academic_research.rst

Lines changed: 5 additions & 6 deletions
Original file line numberDiff line numberDiff line change
@@ -19,20 +19,19 @@ neat-python offers several advantages for academic research:
1919
* **Parallel evaluation** capabilities for computational efficiency
2020
* **Flexible configuration system** allowing systematic parameter exploration
2121
* **Multiple network types** including feedforward, recurrent, CTRNN, and Izhikevich spiking networks
22-
* **Active maintenance** with responsive development and community support
2322
* **Core algorithms correctly implemented** including innovation tracking, crossover, and structural mutations
2423

2524
Relationship to Canonical NEAT
2625
-------------------------------
2726

2827
neat-python correctly implements the most complex and critical aspects of the NEAT algorithm, particularly:
2928

30-
* **Innovation tracking system** with same-generation deduplication and persistence
31-
* **Crossover mechanism** with proper gene alignment by innovation numbers
32-
* **Structural mutations** (add-node and add-connection) following standard NEAT practice
33-
* **75% disable rule** during crossover (with implementation note below)
29+
* **Innovation tracking system** with same-generation deduplication and persistence
30+
* **Crossover mechanism** with proper gene alignment by innovation numbers
31+
* **Structural mutations** (add-node and add-connection) following standard NEAT practice
32+
* **75% disable rule** during crossover (with implementation note below)
3433

35-
However, neat-python contains several implementation choices that differ from the original 2002 NEAT paper. These are not bugs but design decisions that create a NEAT variant with different evolutionary dynamics. Researchers must understand these differences when publishing results.
34+
However, neat-python contains several implementation choices that differ from the original 2002 NEAT paper. These design decisions that create a NEAT variant with different evolutionary dynamics. Researchers must understand these differences when publishing results.
3635

3736
Key Implementation Differences
3837
-------------------------------

0 commit comments

Comments
 (0)