files / problems / hilbert16-limit-cycles / compute / q3 / README.md

1.7 KB · on GitHub

q3 — ten ideas, then the leftover menu and four extras

Imagined end-states, then work backwards. No published H(n) moved.

The claims and the ranking live in ideas.md. Official three plus two extras, then the leftover menu, then four more live lines so workers were not idle:

Line Imagined claim Outcome
ff-two-well/ two-well cubic, 14 Abelian zeros dropped; fork: energy, I(0), L1
gg-pt-lyapunov/ L1 rank 29 at the three PT centers dropped; fork: L1 = 0; two L1(μ)
hh-qh-melnikov/ four zeros of M1 on ẋ=2y, ẏ=−x³ dropped; fork: cyclicity ≤ 1
ii-complex-cube/ holomorphic cube attains 9 sheets dropped; fork: 3-to-1, N=3n+2
jj-weak-hilbert/ beat Z(2,n) by one Abelian zero dropped; fork: radial family attains it
kk-plus-one/ Gasull–Santana +1, two cycles dropped; fork: degree-4 line product, one circle
ll-invariant-line/ cubic + line, three cycles dropped; fork: named cubic, 0 cycles
oo-five-zeros/ five Abelian zeros dropped; fork: 1 on circles, 0 on the cubic nest
pp-christopher-lloyd/ (u²,v²) beats H(7)≥74 dropped; fork: explicit degree-7, four ovals
qq-t2-radial/ T2 radial beats H(7)≥74 dropped; fork: explicit degree-7, four ovals
rr-kolmogorov/ cubic Kolmogorov, 7 cycles dropped; fork: named family, 0 in Q1
ss-cubic-l2/ L2 births extra wells dropped; fork: V2 computed, order is 1
tt-radial-factor/ (x²+y²)X has two cycles dropped; fork: degree 5, still one circle

MM and NN were recycled (same as I and Q) and have no second folder.

Replay:

./run_all.sh

That also runs from the problem compute folder after q1 and q2. Exit 0. Second language: rustc on every line.