Recent activity

2026-09-08 23:56 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17} — length 100000 (was 22572)
2026-09-08 23:38 David Renshaw commented on length 25308 over {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12}
Let A be the 6327-letter {0,...,9} record, word 478, and C(A) its letterwise complement 12-A. Start with A C(A) A C(A), then repair additive squares by weighted conflict search over integer prefix sums, maintaining increments in {0,...,12}. Two rounds with random seeds 60007 and …
2026-09-08 23:38 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12} — length 25308 (was 10776)
2026-09-08 22:50 David Renshaw commented on length 6327 over {0, 1, 2, 3, 4, 5, 6, 7, 8, 9}
Built from the 6316-letter {0,...,10} record (word 436): reverse the word, transform its prefix sums by floor((12*S_i-15*i)/10), and project to increments in {0,...,9} within a minimal uniform band (radius 3). Weighted conflict repair with occasional interval shifts gives a valid…
2026-09-08 22:50 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9} — length 6327 (was 4503)
2026-09-08 22:19 David Renshaw commented on length 100000 over {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21}
Seeded with the first 100,000 letters of the repeating block A C(A), where A is word 476 over {0,...,14} and C replaces each letter x by 21-x. Repaired the additive squares by stochastic local search on integer prefix sums: select a violated three-term progression, change one imp…
2026-09-08 22:19 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20, 21} — length 100000 (was 36538)
2026-09-08 22:04 David Renshaw commented on length 24930 over {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14}
Starting from word 398 over {0,...,19}, subtracted the linear trend 2*i from its prefix sums, then projected the coordinates within distance 7 so all increments lie in {0,...,14}. Repaired additive-square conflicts by changing implicated prefix boundaries, with tabu tenure, occas…
2026-09-08 22:04 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14} — length 24930 (was 14401)
2026-09-08 21:22 David Renshaw commented on length 29320 over {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20}
Compressed the prefix sums of word 281 over {0,...,21} by floor(20*S/21), using its first 29,320 letters. Repaired the resulting additive squares with randomized conflict-guided changes to prefix-sum boundaries, preserving all increment bounds 0..20. The repair selected boundarie…
2026-09-08 21:22 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20} — length 29320 (was 26654)
2026-09-08 17:12 David Renshaw commented on length 3389 over {0, 1, 2, 3, 4, 5, 6, 7, 8}
Improvement of word 445 from 3347 to 3389 over {0,...,8}. Two-hour multistart search using variable-length simulated annealing, exact incremental additive-square counts, shared verified champions, endpoint regrowth, and sum-preserving local rearrangements. Independently checked e…
2026-09-08 17:12 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7, 8} — length 3389 (was 3347)
2026-09-04 22:21 David Renshaw commented on length 2458 over {0, 1, 2, 3, 4, 5, 6, 7}
Improved word 446 from 2,380 to 2,458 letters on {0,...,7}. Alternated finite-domain SAT suffix rewrites, endpoint reversal, mobility-guided iterated DFS, and guide-preserving rollout beams. Also developed an incremental PicoSAT solver with cumulative-sum variables, digit channel…
2026-09-04 22:21 David Renshaw extended {0, 1, 2, 3, 4, 5, 6, 7} — length 2458 (was 2380)
2026-08-31 01:33 Aayush Rajasekaran commented on length 215 over {0, 1, 4, 8}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 216 exists over this alphabet, so max =…
2026-08-31 01:33 Aayush Rajasekaran extended {0, 1, 4, 8} — length 215 (was 205)
2026-08-31 00:40 Aayush Rajasekaran commented on length 200 over {0, 1, 2, 8}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 201 exists over this alphabet, so max =…
2026-08-31 00:40 Aayush Rajasekaran extended {0, 1, 2, 8} — length 200 (was 179)
2026-08-31 00:06 Aayush Rajasekaran commented on length 195 over {0, 2, 5, 8}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 196 exists over this alphabet, so max =…
2026-08-31 00:06 Aayush Rajasekaran extended {0, 2, 5, 8} — length 195 (was 193)
2026-08-30 23:36 Aayush Rajasekaran commented on length 184 over {0, 2, 3, 8}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 185 exists over this alphabet, so max =…
2026-08-30 23:36 Aayush Rajasekaran extended {0, 2, 3, 8} — length 184 (was 160)
2026-08-30 22:05 Aayush Rajasekaran commented on length 167 over {0, 1, 5, 7}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 168 exists over this alphabet, so max =…
2026-08-30 22:05 Aayush Rajasekaran extended {0, 1, 5, 7} — length 167 (was 163)
2026-08-30 19:57 Aayush Rajasekaran commented on length 148 over {0, 1, 2, 7}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 149 exists over this alphabet, so max =…
2026-08-30 19:57 Aayush Rajasekaran extended {0, 1, 2, 7} — length 148 (was 147)
2026-08-30 19:40 Aayush Rajasekaran commented on length 139 over {0, 2, 3, 7}
This word attains the exact maximum: exhaustive search (Rust backtracking engine over prefix sums; counts cross-validated against a naive implementation and against this site's closed alphabets) proves no additive-square-free word of length 140 exists over this alphabet, so max =…
2026-08-30 19:40 Aayush Rajasekaran extended {0, 2, 3, 7} — length 139 (was 137)
2026-08-30 19:36 Aayush Rajasekaran commented on length 205 over {0, 1, 4, 8}
Found by parallel randomized DFS with frontier-biased restarts. Our exhaustive search additionally proves the exact maximum over this alphabet is 215 (closure claim to follow); this witness has length 205.