Skip to content

Commit f1dd2fc

Browse files
committed
Correct three bound-table values and drop an impossible upper bound
Recomputed every closed-form/decimal pair recorded in constants/ and README.md. Three recorded values disagree with their own closed form or with the source they cite: - 1a: the Schinzel-Schmidt row prints $\pi/2 = 1.57059$, but $\pi/2 = 1.5707963\dots$. - 1b: the Swierczkowski row prints $(4-\sqrt{6})/5 \approx 0.310679$, but $(4-\sqrt{6})/5 = 0.3101020\dots$. - 1b: the Haugland 2016 row prints 0.380927; the abstract of arXiv:1609.08000 gives 0.380926. 8a also carries $0$ as an upper bound on C_8. That asserts C_8 <= 0, contradicting the lower bound 2/log(gamma_1) = 0.755106 recorded on the same page, and every later row in the same table is larger, so it cannot be a chronological first entry either. Removed it, and recorded on the lower-bound row why 2/log(gamma_1) holds unconditionally (the underscores there are escaped for kramdown).
1 parent 3a14910 commit f1dd2fc

3 files changed

Lines changed: 4 additions & 5 deletions

File tree

‎constants/1a.md‎

Lines changed: 1 addition & 1 deletion
Original file line numberDiff line numberDiff line change
@@ -12,7 +12,7 @@ for all non-negative $f \colon \mathbb{R} \to \mathbb{R}$.
1212

1313
| Bound | Reference | Comments |
1414
| ----- | --------- | -------- |
15-
| $\pi/2 = 1.57059$ | [SS2002] | |
15+
| $\pi/2 = 1.5707963\dots$ | [SS2002] | |
1616
| $1.50992$ | [MV2009] | |
1717
| $1.5053$ | [GGSWT2025] | May 2025 announcement, AlphaEvolve
1818
| $1.503164$ | [GGSWT2025] | Dec 2025 preprint release, AlphaEvolve

‎constants/1b.md‎

Lines changed: 2 additions & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -19,7 +19,7 @@ for all non-negative $f,g: [-1,1] \to [0,1]$ with $f+g=1$ on $[-1,1]$ and $\int_
1919
| $0.4$ | [MRS1956]| |
2020
| $0.385694$ | Haugland (unpublished, 1993) | |
2121
| $0.382002$ | [H1996] | |
22-
| $0.380927$ | [H2016] | |
22+
| $0.380926$ | [H2016] | |
2323
| $0.380924$ | [GGSWT2025] | AlphaEvolve |
2424
| $0.380876$ | [YKLBMWKCZGS2026] | TTT-Discover |
2525
| $0.380871$ | [T2026] | TogetherAI |
@@ -31,7 +31,7 @@ for all non-negative $f,g: [-1,1] \to [0,1]$ with $f+g=1$ on $[-1,1]$ and $\int_
3131
| ----- | --------- | -------- |
3232
| $1/4=0.25$ | [E1955] | |
3333
| $1-1/\sqrt{2} \approx 0.292893$ | Scherk (unpublished, 1955) | |
34-
| $(4-\sqrt{6})/5 \approx 0.310679$ | [S1958] | |
34+
| $(4-\sqrt{6})/5 \approx 0.310102$ | [S1958] | |
3535
| $\sqrt{4-\sqrt{15}} \approx 0.356393$ | [M1959] | |
3636
| $0.379005$ | [W2022] |
3737

‎constants/8a.md‎

Lines changed: 1 addition & 2 deletions
Original file line numberDiff line numberDiff line change
@@ -8,7 +8,6 @@ $C\_{8} = R$ is the least constant such that there are no zeroes $\sigma+it$ of
88

99
| Bound | Reference | Comments |
1010
| ----- | --------- | -------- |
11-
| $0$ | Trivial | |
1211
| $34.82$ | [dlVP1899] | Implies the prime number theorem|
1312
| $19$ | [RS1962] | |
1413
| $9.64591$ | [S1970] | |
@@ -23,7 +22,7 @@ $C\_{8} = R$ is the least constant such that there are no zeroes $\sigma+it$ of
2322

2423
| Bound | Reference | Comments |
2524
| ----- | --------- | -------- |
26-
| $2/\log \gamma_{1} \approx 0.755106$ | - | Optimal assuming RH |
25+
| $2/\log \gamma\_{1} \approx 0.755106$ | - | Unconditional: $\gamma\_{1} = 14.1347\dots$ is the ordinate of the lowest zero of $\zeta$, and that zero has $\sigma = 1/2$, so any admissible $R$ must satisfy $1/2 \leq 1 - 1/(R\log\gamma\_{1})$. Optimal assuming RH. |
2726

2827
## Additional comments and links
2928

0 commit comments

Comments
 (0)