9x9 Column Proposals
Summary
- No 9x9 square yet, and no partial reaches seven rows without junk column strings. The cleanest wide scan has 253 six-row partials. Adding the whole judged fill tier gives 430, none at seven.
- Recommendation: stop adding words at these top rows and test an 8×9 rectangle next. Each further row has cost a factor of about 4,500 to 7,000, and three more rows are needed.
- Columns are what stops the search. Allowing corpus strings in the columns gets it to seven rows, with real rows.
- Those strings are mostly junk or names, and a partial needs several at once. Adding everything gemma4 and Gemini accepted gives 70 six-row and 1 seven-row partials.
Clean scan: first real six-row partials
Without the fill tier, every row and column comes from cat1 plus words I reviewed. The search is then small enough to nearly finish. With rows 1–3 limited to common words, it covers all 5,375 start words in about two minutes, and 24 are cut by budget.
| Scratch lexicon | Top rows | Five-row | Six-row | Seven-row | Squares |
|---|---|---|---|---|---|
| 1,128 words | common | 76,300 | 8 | 0 | 0 |
| + 175 words | common | 76,995 | 26 | 0 | 0 |
| + 9 words | common | 76,998 | 35 | 0 | 0 |
| + 20 reviewed fill words | common | 77,073 | 35 | 0 | 0 |
| + 65 words (1,397 in all) | common | 77,186 | 35 | 0 | 0 |
| Before those 65 (1,332 words) | wider | 875,103 | 116 | 0 | 0 |
| + 65 words (1,397 in all) | wider | 876,502 | 200 | 0 | 0 |
| Larger reviewed lexicon (lex12) | wider | 1,112,411 | 252 | 0 | 0 |
| + 159 unaimed field words (lex13) | wider | 1,129,459 | 253 | 0 | 0 |
| Judged fill tier (lex15, 11,238 words) | common | 241,678 | 52 | 0 | 0 |
| Judged fill tier (lex15, 11,238 words) | wider | 3,078,459 | 430 | 0 | 0 |
Wider means 13,193 start words. Cuts by budget vary by run: about 1,890 in the first wider runs, 5 in lex13, 1,635 of 13,193 in the wider lex15 run, so that row is a sample and is not directly comparable with lex13. The fill-tier rows use strings accepted by judges, most of them not real words by my reading. The other added words are my reading, not graded by Eric.
One example: sharpness / tolerance / rotterdam / overscore / merosomes / brightest, which relies on my pressheet.
Row 7 is the new wall. Of 123 six-row partials checked in the wider scan, only 2 have a listed row word within two column misses. No real word is in sight for prefixes such as tinamib or flaptrf.
What free columns buy
This section and the next use the older setup with the fill tier. I took the 3,272 five-row partials from its last scan and resumed them.
Rows came from the lexicon. Columns could also use strings from the 9-letter corpus pool, which is full of junk, so the two pool rows of this table are a ceiling.
| Columns may use | Six-row | Seven-row | Eight-row | Squares |
|---|---|---|---|---|
| Lexicon only | 0 | 0 | 0 | 0 |
| + pool, count ≥ 100 (1.36M) | 8,927 | 13 | 0 | 0 |
| + all pool (2.48M) | 59,942 | 277 | 0 | 0 |
| + 71 real words I picked | 0 | 0 | 0 | 0 |
| + those 71 and 687 judge-accepted strings | 70 | 1 | 0 | 0 |
Fresh 48-start scans: with the 71 words, 3,255 lines at five rows and 0 at six. With the judged strings too, 3,305 at five, 68 at six, 1 at seven. The base was 3,243 at five. All starts were cut by budget, so these are samples. Candidates were chosen because they fit these partials, so some of the gain is by construction.
Why real words fall short
One of the easiest six-row partials. It needs two new column words, starting restop and sealss. The first could be real; the second is hard to believe.
Few partials are this easy, and each new word also pins the letters of the rows below it.
| New column words needed (pool count ≥ 100) | Six-row partials |
|---|---|
| 1 | 1 |
| 2 | 34 |
| 3 | 211 |
| 4 or more | 8,669 |
Counted from up to three partials stored per pattern. Every seven-row partial in this group needs five or more.
The strings these partials lean on are mostly junk or names: seanessie, stiverson, sellstrom. A few are real: carnaroli, seastacks, stoplists.
llama3.1:8b could not propose from prefixes: 220 prefixes gave 0 usable words before the run was stopped.
The six-row partials closest to real, by my reading:
- propagate / retrieved / ephedrine / chessmail / irrigates / naturtint needs one new column starting othert, such as otherthan. Its rows include chessmail and naturtint.
- saturated / oppressor / aerospace / prospered / sistering / cotelette needs soapscale and asperezas. The second isn't English.
The one seven-row partial needs at least appalatus or appalatia, saliaceae and seaness…, none of them real words.
Judges
Scored from verdicts already on file. Each cell is strings accepted. A second judge sees only what gemma4 rejected, about 29% of strings.
| Accepted | gemma4 alone | + gemini-2.5-flash | + mistral-large-2512 |
|---|---|---|---|
| Real words (of 200) | 172 | 184 | 182 |
| Junk (of 72) | 22 | 27 | 22 |
| Eric ≥ 3 (of 142) | 112 | 128 | 119 |
| Eric < 3 (of 39) | 14 | 24 | 20 |
On the 1,675 column candidates, gemma4 accepted 544 and took 610 s. Gemini then accepted 143 of the 1,131 strings left, for $0.04 (46 calls, none failed). Both lists hold much junk, such as othermeat and seanessie. The test sets above were cleaner than these candidates, so don't expect those rates here.
Mistral failure
mistral-large-4-0 now refuses requests with reasoning off (HTTP 400, code 3051), which is what word_oracle sends. With reasoning on it answers, at $1.14 per 1,000 strings and two minutes per batch. word_oracle now uses mistral-large-2512 as its second judge. A single probe call passed; it has not been run end to end.
Limits of these results
- The words I added are my reading of candidates. They and the judged strings have no grades from Eric.
- Resumes start only from partials the lexicon already reached. A new word could open different upper rows, which only the 48-start scan samples.
- The pool is a corpus, so these counts do not show that a square can or cannot exist.
Decisions needed
- Chris: stop adding words at these top rows and try an 8×9 rectangle? Five-to-six-row rates are 1 in 4,464 (lex13) and 1 in 7,159 (lex15). If later rows cost the same, which is only a rough assumption, a square would need about a billion times more six-row partials than the 430 found.
- Why a rectangle: it reuses the same row words, and 8-letter columns have far more words to draw on than 9-letter ones. Its record status needs checking.
- Eric: grades for the sample words, sent separately.