analiza aiqbench
This research advances machine learning by demonstrating how bio-inspired meta-learning strategies can optimize activation function discovery in indirectly encoded neural networks. The study highlights that matching the timescale of biological adaptation mechanisms to evolutionary evaluation windows significantly improves convergence speed and solve rates for complex problems like parity. For AI practitioners, the key takeaway is that curated function palettes guided by specific temporal dynamics outperform both standard baselines and exhaustive search methods.
Bio-Inspired Palette Evolution in Indirectly Encoded Substrates: Timescale Compatibility Shapes Activation Function Discovery
arXiv:2609.17067v1 Announce Type: cross Abstract: Indirectly encoded neural networks can assign different activation functions to individual nodes, but the right functions are rarely known in advance. When the available set contains only standard monotonic functions, problems like parity become unsolvable, yet an all-inclusive palette underperforms a curated one. How should evolution discover which functions to use? We address this as a meta-learning problem, designing 13 strategies (11 inspired by biological adaptation mechanisms, plus baseline and oracle controls) that modify the set of available activation functions during evolution. Each strategy translates a biological principle into an evolutionary operator: for example, circadian-inspired oscillatory gating cycles functions in and out of the palette on a fixed schedule, while immune-inspired Clonal Selection permanently protects functions that consistently correlate with fitness. We evaluate all strategies across more than 3,000 runs on parity and non-parity problems, first evolving the activation palette alone, then co-evolving a per-node aggregation palette on harder problems; an independent replication with new seeds confirms a stable high-reliability tier, with Circadian holding its top rank. Bio-inspired strategies match the solve rate of a tuned baseline but converge up to twice as fast, with Circadian halving total compute. Strategy rankings reverse across problem types, with no strategy dominating all domains. Strategy success is largely shaped by timescale compatibility: strategies whose characteristic timescale matches the evolutionary evaluation window consistently outperform those that operate too slowly. The practical guideline: match the mechanism's timescale to the evaluation budget. Rescaling the slowest strategy bypasses the oscillatory barrier entirely: all nine solutions solve parity with non-oscillatory activations paired with min or max aggregation.
Abstract:Indirectly encoded neural networks can assign different activation functions to individual nodes, but the right functions are rarely known in advance. When the available set contains only standard monotonic functions, problems like parity become unsolvable, yet an all-inclusive palette underperforms a curated one. How should evolution discover which functions to use? We address this as a meta-learning problem, designing 13 strategies (11 inspired by biological adaptation mechanisms, plus baseline and oracle controls) that modify the set of available activation functions during evolution. Each strategy translates a biological principle into an evolutionary operator: for example, circadian-inspired oscillatory gating cycles functions in and out of the palette on a fixed schedule, while immune-inspired Clonal Selection permanently protects functions that consistently correlate with fitness. We evaluate all strategies across more than 3,000 runs on parity and non-parity problems, first evolving the activation palette alone, then co-evolving a per-node aggregation palette on harder problems; an independent replication with new seeds confirms a stable high-reliability tier, with Circadian holding its top rank. Bio-inspired strategies match the solve rate of a tuned baseline but converge up to twice as fast, with Circadian halving total compute. Strategy rankings reverse across problem types, with no strategy dominating all domains. Strategy success is largely shaped by timescale compatibility: strategies whose characteristic timescale matches the evolutionary evaluation window consistently outperform those that operate too slowly. The practical guideline: match the mechanism's timescale to the evaluation budget. Rescaling the slowest strategy bypasses the oscillatory barrier entirely: all nine solutions solve parity with non-oscillatory activations paired with min or max aggregation.
16 pages, 2 figures, 7 tables. Authors' accepted manuscript; published in Parallel Problem Solving from Nature - PPSN XIX (Springer, Lecture Notes in Computer Science)
Neural and Evolutionary Computing (cs.NE); Machine Learning (cs.LG)
arXiv:2609.17067 [cs.NE]
(or arXiv:2609.17067v1 [cs.NE] for this version)
https://doi.org/10.48550/arXiv.2609.17067 arXiv-issued DOI via DataCite (pending registration)
Parallel Problem Solving from Nature - PPSN XIX, Lecture Notes in Computer Science, Springer Nature Switzerland, Cham, 2026, pp. 368-383
https://doi.org/10.1007/978-3-032-36217-9_23 DOI(s) linking to related resources
Text integral sub licența deschisă a publicației · 333 cuvinte
Subiecte conexe
Does preservation make sense before we know how to revive?
LessWrong · 15 iun. 2026
There Is No Alignment Without Value Stability
LessWrong · 17 sept. 2026
Per-Node Activation Function Evolution in Indirectly Encoded Substrates: Solvability, Limits, and Emergent Diversity
arXiv · 2 oct. 2026
Breaking the Central Bias: Spatially Partitioned Experts for Coordinate-Based Neuroevolution
arXiv · 11 sept. 2026