Evolution
AIGARTH-EVO-01 — the ternary rule-table substrate. 10 / 10 environments converged · 3,712 total evaluations · last run 0ms.
AIGARTH-EVO-01 battery
Ten environments from the substrate — boolean logic (AND, OR, NAND, XOR, MAJ3, PARITY3), sequence prediction (COPY_K, NOT_K, XOR_K), and BTC direction. Each is a swappable fitness function over the same ternary rule-table engine.
and
converged @ gen 3
trinary-rule-table@1
best fitness
100.0%
generations
4
evaluations
256
nodes
8(2 in / 1 out)
v38a8b4ce7-3
bestmean
or
converged @ gen 2
trinary-rule-table@1
best fitness
100.0%
generations
3
evaluations
192
nodes
8(2 in / 1 out)
v4bc476bfc-4
bestmean
nand
converged @ gen 0
trinary-rule-table@1
best fitness
100.0%
generations
1
evaluations
64
nodes
8(2 in / 1 out)
v0d8175675-0
bestmean
xor
converged @ gen 5
trinary-rule-table@1
best fitness
100.0%
generations
6
evaluations
384
nodes
8(2 in / 1 out)
v7916a872a-7
bestmean
maj3
converged @ gen 7
trinary-rule-table@1
best fitness
100.0%
generations
8
evaluations
512
nodes
8(3 in / 1 out)
v6e341af95-6
bestmean
parity3
converged @ gen 31
trinary-rule-table@1
best fitness
100.0%
generations
32
evaluations
2,048
nodes
12(3 in / 1 out)
v8c3500e4f-8
bestmean
copy_k
converged @ gen 0
trinary-rule-table@1
best fitness
100.0%
generations
1
evaluations
64
nodes
16(4 in / 1 out)
v06813cbda-0
bestmean
not_k
converged @ gen 0
trinary-rule-table@1
best fitness
100.0%
generations
1
evaluations
64
nodes
16(4 in / 1 out)
v017708572-0
bestmean
xor_k
converged @ gen 0
trinary-rule-table@1
best fitness
100.0%
generations
1
evaluations
64
nodes
16(4 in / 1 out)
v0eb610c12-0
bestmean
btc_direction
converged @ gen 0
trinary-rule-table@1
best fitness
100.0%
generations
1
evaluations
64
nodes
16(5 in / 1 out)
v09c69ac4f-0
bestmean
Global Contribution Pool
Synthetic (Work Runtime unreachable)
last hour
14.8K
evaluations
last day
287.4K
evaluations
running
9
work items
queued
14
work items
Workers
7
light (browser)
3
dedicated
local
2 active/ 4
remote
1 active/ 2
browser
7 active/ 12
oc-processor
0 active/ 1
neuraxon
0 active/ 0
human-via-oracle
0 active/ 0
Evaluations by environment (last 24h)
evo_step192.0K (66.8%)
awork_173.4K (25.5%)
btc_direction_predictor22.0K (7.7%)
What you're looking at
AIGARTH-EVO-02 — global contribution pool.
The substrate is now executable in your browser via WebAssembly. Toggling Contribute CPU spawns a Web Worker that runs the same deterministic evaluation function as the server, on this tab. The pool widget aggregates the live state of every worker kind.
Light workers = browser tabs. Dedicated workers = local and remote server-class. The bar is the live ratio; the rows below break it down by kind.
Your share of the pool is computed live asyour_evals / global_evals_last_hour. Browser workers start at reputation 0.3 (vs 0.5 for local); the Work Runtime verifier spot-checks every browser submission against a server re-run.
Wire: the
evo_step algorithm is registered in services/work (port 7012) alongsideawork_1. This dashboard consumes @aigarth/evo directly; the worker container is the next-step delivery.