gpt-5 v3 Evaluation

Generated: 2026-09-20 11:21:08 UTC

Pairs100
Distinct lemmas100
Slope0.902
R^20.986
BLEU-449.1%
chrF++72.3%
METEOR78.3%
ROUGE-L79.3%
BERTScore96.2%
COMET79.9%
BLEURT75.2%

Profile version id: 2342

First translation using this version: 2026-07-15

Total usable AI runs for this version: 100 runs across 100 lemmas

Created: 2026-07-15 13:49:25.227272+10:00

Active: True

Notes: Model timeline evaluation: GPT-5 using prompt text from gpt-5.5 v3; approved-human-only for the 100-row Kappa corpus.

Prompt length scatter plot

Length Regression

FormulaAI words = intercept + slope * human words
Slope0.902
Intercept2.087
R^20.986
P-value2.47e-92
Pearson r0.993
Mean source words30.400
Length fallback rows0
Mean human words44.150
Mean AI words41.920
Mean absolute length residual2.870
Mean absolute percent error7.1%

Metric vs Passage Length

These plots test whether score changes with source passage length for this prompt version. The dashed line is an ordinary least-squares fit; the annotation reports n, r, R^2, p-value, and slope.

BLEU-4 vs passage length
BLEU-4: significant negative; r = -0.374, R^2 = 0.140, p = 0.0001.
chrF++ vs passage length
chrF++: significant negative; r = -0.443, R^2 = 0.196, p = 3.95e-06.
METEOR vs passage length
METEOR: significant negative; r = -0.524, R^2 = 0.274, p = 2.23e-08.
ROUGE-L vs passage length
ROUGE-L: significant negative; r = -0.471, R^2 = 0.221, p = 7.83e-07.
BERTScore vs passage length
BERTScore: significant negative; r = -0.610, R^2 = 0.372, p = 1.64e-11.
COMET vs passage length
COMET: significant negative; r = -0.479, R^2 = 0.229, p = 4.69e-07.
BLEURT vs passage length
BLEURT: significant negative; r = -0.753, R^2 = 0.567, p = 1.65e-19.
Trigram precision vs passage length
Trigram precision: significant negative; r = -0.370, R^2 = 0.137, p = 0.0002.
Trigram recall vs passage length
Trigram recall: significant negative; r = -0.378, R^2 = 0.143, p = 0.0001.
Trigram F1 vs passage length
Trigram F1: significant negative; r = -0.374, R^2 = 0.140, p = 0.0001.
Trigram Jaccard vs passage length
Trigram Jaccard: significant negative; r = -0.368, R^2 = 0.135, p = 0.0002.
Metric Rows Pattern Pearson r R^2 P-value Slope Status
BLEU-4 100 significant negative -0.374 0.140 0.0001 -0.00238 ok
chrF++ 100 significant negative -0.443 0.196 3.95e-06 -0.00165 ok
METEOR 100 significant negative -0.524 0.274 2.23e-08 -0.00248 ok
ROUGE-L 100 significant negative -0.471 0.221 7.83e-07 -0.00197 ok
BERTScore 100 significant negative -0.610 0.372 1.64e-11 -0.00042 ok
COMET 100 significant negative -0.479 0.229 4.69e-07 -0.00089 ok
BLEURT 100 significant negative -0.753 0.567 1.65e-19 -0.00259 ok
Trigram precision 100 significant negative -0.370 0.137 0.0002 -0.00305 ok
Trigram recall 100 significant negative -0.378 0.143 0.0001 -0.00312 ok
Trigram F1 100 significant negative -0.374 0.140 0.0001 -0.00308 ok
Trigram Jaccard 100 significant negative -0.368 0.135 0.0002 -0.00336 ok

Translation Similarity Metrics

Mean BLEU-449.1%
Mean chrF++72.3%
Mean METEOR78.3%
Mean ROUGE-L79.3%
Mean BERTScore96.2%
Mean COMET79.9%
Mean BLEURT75.2%
Legacy smoothed corpus BLEU50.0%
Unigram F179.5%
Bigram F157.0%
Trigram precision45.2%
Trigram recall42.8%
Trigram F143.9%
Trigram Jaccard28.2%
4-gram F134.7%

Largest Length Residuals

Lemma ID Headword Human words AI words Residual BLEU-4 chrF++ METEOR ROUGE-L BERTScore COMET BLEURT Human Run Model
2484 Καρία 270 227 -18.687 30.6% 57.0% 54.5% 64.4% 91.5% 79.5% 55.5% reviewed/approved 11782 gpt-5-2025-08-07
2603 Κάρυστος 175 178 18.024 40.6% 66.1% 62.7% 66.3% 94.8% 78.4% 57.7% reviewed/approved 11794 gpt-5-2025-08-07
7254 Κύτα 84 88 10.126 39.2% 64.2% 67.2% 69.8% 94.4% 77.7% 63.2% reviewed/approved 11893 gpt-5-2025-08-07
2470 Καππαδοκία 68 72 8.562 26.0% 64.1% 81.7% 77.1% 95.7% 75.2% 64.6% reviewed/approved 11776 gpt-5-2025-08-07
2055 Καβασσός 109 92 -8.429 27.9% 58.8% 55.9% 57.7% 93.4% 76.8% 50.9% reviewed/approved 11638 gpt-5-2025-08-07
2604 Καρχηδών 117 116 8.353 39.4% 66.2% 70.0% 71.5% 94.0% 78.9% 63.9% reviewed/approved 11797 gpt-5-2025-08-07
2342 Κάμιρος 54 59 8.193 51.0% 76.2% 76.0% 79.6% 97.3% 80.2% 73.2% reviewed/approved 11758 gpt-5-2025-08-07
7251 Κύρτος 66 54 -7.634 40.1% 62.1% 61.1% 72.7% 94.1% 71.9% 64.3% reviewed/approved 11884 gpt-5-2025-08-07
2328 Καλλίπολις 43 48 7.118 35.2% 64.8% 73.0% 65.9% 93.6% 71.4% 66.0% reviewed/approved 11734 gpt-5-2025-08-07
7255 Κυτέριον 22 29 7.064 43.8% 70.2% 73.2% 70.6% 96.0% 77.3% 76.1% reviewed/approved 11896 gpt-5-2025-08-07
7260 Κυχρεῖος πάγος 61 64 6.878 55.7% 73.9% 80.3% 75.2% 96.1% 77.1% 64.5% reviewed/approved 11911 gpt-5-2025-08-07
2607 Κάσος 57 60 6.486 44.0% 70.9% 79.6% 83.8% 95.9% 76.4% 74.4% reviewed/approved 11803 gpt-5-2025-08-07
2625 Κατακεκαυμένη 46 50 6.411 56.0% 76.0% 85.7% 79.2% 96.5% 78.2% 76.5% reviewed/approved 11824 gpt-5-2025-08-07
2056 Καβειρία 112 97 -6.136 36.6% 69.0% 61.1% 67.9% 95.5% 78.8% 60.3% reviewed/approved 11641 gpt-5-2025-08-07
7257 Κυτώνιον 25 19 -5.642 39.8% 66.7% 75.0% 81.8% 95.6% 79.2% 74.2% reviewed/approved 11902 gpt-5-2025-08-07
2468 Καπετώλιον 138 132 5.407 28.6% 56.1% 56.8% 64.4% 89.7% 70.6% 54.7% reviewed/approved 11773 gpt-5-2025-08-07
2621 Κάστνιον 25 30 5.358 34.3% 69.8% 80.0% 69.1% 96.5% 80.8% 73.2% reviewed/approved 11812 gpt-5-2025-08-07
2119 Κάλλατις 65 65 4.269 36.5% 62.5% 71.6% 70.8% 95.3% 69.7% 65.6% reviewed/approved 11722 gpt-5-2025-08-07
2455 Κάναστρον 60 52 -4.220 22.3% 57.6% 45.8% 55.4% 93.8% 71.1% 58.6% reviewed/approved 11767 gpt-5-2025-08-07
2597 Καρπασία 112 99 -4.136 41.9% 69.1% 61.2% 65.4% 94.3% 74.1% 55.4% reviewed/approved 11785 gpt-5-2025-08-07
2465 Κάνωπος 83 81 4.029 39.6% 66.6% 71.5% 76.8% 95.8% 77.7% 70.1% reviewed/approved 11770 gpt-5-2025-08-07
2059 Καδμεία 20 24 3.869 31.4% 70.5% 82.7% 81.8% 96.6% 84.3% 81.7% reviewed/approved 11650 gpt-5-2025-08-07
2335 Κάλυτις 26 22 -3.545 39.2% 65.2% 68.8% 75.0% 96.9% 81.1% 79.8% reviewed/approved 11755 gpt-5-2025-08-07
2329 Κάλπη 57 50 -3.514 49.0% 69.4% 69.1% 79.6% 96.4% 76.0% 72.1% reviewed/approved 11737 gpt-5-2025-08-07
2605 Κάσιον 56 56 3.389 51.6% 72.7% 74.7% 76.8% 96.2% 76.0% 67.9% reviewed/approved 11800 gpt-5-2025-08-07
2080 Καλαβρία 18 15 -3.327 36.9% 64.0% 61.1% 66.7% 94.0% 80.8% 75.6% reviewed/approved 11683 gpt-5-2025-08-07
2114 Κάλβιος 18 15 -3.327 20.7% 49.8% 59.2% 60.6% 94.0% 72.4% 70.0% reviewed/approved 11707 gpt-5-2025-08-07
2081 Καλάθη 49 43 -3.296 35.0% 67.4% 76.5% 82.6% 96.0% 74.7% 72.3% reviewed/approved 11686 gpt-5-2025-08-07
2079 Καισάρεια 39 34 -3.274 57.9% 75.6% 80.4% 84.9% 97.8% 82.0% 79.7% reviewed/approved 11680 gpt-5-2025-08-07
7259 Κύφος 54 54 3.193 49.1% 73.3% 83.4% 81.5% 96.5% 75.9% 73.1% reviewed/approved 11908 gpt-5-2025-08-07

Residual Predictors

Positive signed-residual terms are associated with AI translations longer than expected from the human length; negative signed-residual terms are associated with shorter AI translations. Absolute-residual models identify terms associated with larger AI-human length divergence.

Greek source terms predicting signed length residuals

Features: 176; ridge alpha: 2.031; cross-validated R^2: -0.093.

Positive termCoefficientDocsMean presentMean absent
οι 2.8070 11 4.558 -0.563
ος 2.6667 6 6.530 -0.417
τη 2.1701 15 2.300 -0.406
εις 2.0294 7 3.275 -0.247
εν τη 1.9746 7 3.906 -0.294
ου και 1.9580 4 9.224 -0.384
και το 1.6815 9 3.335 -0.330
λεγεται και 1.6749 8 3.575 -0.311
λεγεται 1.6601 9 3.191 -0.316
απο 1.6567 27 1.664 -0.615
εξ 1.6028 3 6.964 -0.215
εκληθη 1.4617 4 7.078 -0.295
υπο 1.4058 4 6.051 -0.252
και 1.3792 66 0.279 -0.542
δε και 1.3775 8 1.707 -0.148
ορος 1.3716 6 5.317 -0.339
τε 1.3624 5 3.110 -0.164
ωστε 1.3450 3 8.263 -0.256
ου 1.2811 14 1.815 -0.296
στραβων 1.2169 15 2.505 -0.442
Negative termCoefficientDocsMean presentMean absent
καππαδοκιας -1.9597 3 -4.481 0.139
οικητωρ -1.8836 6 -5.031 0.321
μεταξυ -1.5815 5 -3.050 0.161
ως το -1.5626 3 -7.399 0.229
θεοπομπος -1.4753 5 -3.252 0.171
φησιν -1.3813 10 -1.807 0.201
το θηλυκον -1.2024 6 -2.476 0.158
μεν -1.2004 4 -7.316 0.305
το -1.1843 70 -0.013 0.030
ελλανικος -1.1132 4 -7.436 0.310
παρα -1.0961 7 -3.741 0.282
ει -1.0640 4 -5.000 0.208
λυκιας -1.0570 3 -4.499 0.139
τω περι -1.0499 4 -5.157 0.215
εν τω περι -1.0499 4 -5.157 0.215
ως -1.0337 52 -0.336 0.364
εν τω -1.0305 6 -0.756 0.048
το κτητικον -0.9929 7 -3.379 0.254
κτητικον -0.9929 7 -3.379 0.254
και του -0.9850 4 -3.644 0.152

Greek source terms predicting large absolute residuals

Features: 176; ridge alpha: 2.894; cross-validated R^2: -0.201.

Positive termCoefficientDocsMean presentMean absent
τω 2.0205 16 5.644 2.341
και 1.9521 66 3.518 1.612
δε 1.6939 46 4.172 1.760
της 1.5771 35 4.685 1.892
την 1.5271 13 6.863 2.273
τη 1.4779 15 5.652 2.379
εις 1.3824 7 9.236 2.391
γαρ 1.2386 9 7.539 2.408
εν τω 1.1350 6 9.137 2.470
λεγεται και 1.1119 8 8.588 2.373
εκαλειτο δε 1.1078 3 15.021 2.494
απο 1.1033 27 4.935 2.106
δια 1.0788 16 5.171 2.431
απο της 1.0050 6 8.907 2.484
φησι 0.9612 12 6.374 2.392
εκαλειτο δε και 0.9478 3 15.021 2.494
εκαλειτο 0.9388 7 8.857 2.419
αφ 0.9315 6 6.637 2.629
εν 0.9272 37 4.078 2.160
λεγεται 0.8962 9 7.647 2.397
Negative termCoefficientDocsMean presentMean absent
το εθνικον -0.8732 56 2.894 2.839
εθνικον -0.8732 56 2.894 2.839
προς -0.8268 15 2.773 2.887
κωμη -0.6924 3 1.103 2.924
πλησιον -0.6915 8 1.479 2.991
πολις -0.6770 68 2.988 2.618
προς τη -0.6723 4 1.427 2.930
τους -0.5320 4 1.601 2.923
στραβων -0.5058 15 3.557 2.749
και πολις -0.4872 5 2.874 2.870
πολιτης -0.4793 18 3.191 2.799
νησος -0.4677 9 2.606 2.896
πολυβιος -0.4403 5 1.516 2.941
τεταρτω -0.4021 3 1.349 2.917
εκαταιος ασια -0.3927 5 0.989 2.969
ασια -0.3927 5 0.989 2.969
πορρω -0.3750 3 1.174 2.922
ου πορρω -0.3750 3 1.174 2.922
ακρα -0.3094 4 2.250 2.896
φησι το -0.3033 3 1.378 2.916

AI English terms predicting large absolute residuals

Features: 312; ridge alpha: 4.125; cross-validated R^2: -0.315.

Positive termCoefficientDocsMean presentMean absent
the 1.7037 94 2.955 1.536
and 1.1577 56 3.803 1.682
also 1.1419 43 4.198 1.868
that 1.0287 12 6.761 2.339
for the 0.8764 7 7.781 2.500
to 0.8731 21 5.116 2.273
for 0.8584 14 6.114 2.342
used 0.8500 10 8.097 2.289
after 0.8334 22 4.615 2.378
to the 0.8209 5 10.311 2.478
is 0.7807 90 3.010 1.606
of 0.6985 63 3.463 1.859
with 0.6969 18 4.077 2.605
usage 0.6852 4 7.366 2.682
was 0.6562 16 5.706 2.329
there 0.6447 32 4.306 2.194
diphthong 0.6426 3 9.410 2.667
after the 0.6410 4 12.073 2.486
says 0.6333 16 5.499 2.369
is also 0.5971 31 3.995 2.364
Negative termCoefficientDocsMean presentMean absent
near -0.5450 19 2.621 2.928
city in -0.5265 41 2.512 3.119
the ethnonym is -0.5125 54 3.019 2.694
ethnonym is -0.5125 54 3.019 2.694
a city in -0.4888 41 2.512 3.119
as in -0.4280 21 2.942 2.850
city of the -0.4089 10 2.040 2.962
the ethnonym -0.4037 57 3.086 2.583
ethnonym -0.4037 57 3.086 2.583
a city -0.3774 65 3.062 2.512
city of -0.3605 13 2.516 2.923
city near -0.3031 3 0.962 2.929
a city near -0.3031 3 0.962 2.929
say -0.2974 9 2.501 2.906
in his asia -0.2968 5 0.989 2.969
his asia -0.2968 5 0.989 2.969
a city of -0.2883 11 2.664 2.895
book -0.2754 39 2.985 2.796
city -0.2478 68 2.988 2.618
asia -0.2340 6 1.847 2.935

Human English terms predicting large absolute residuals

Features: 335; ridge alpha: 4.125; cross-validated R^2: -0.310.

Positive termCoefficientDocsMean presentMean absent
the 1.2987 90 3.013 1.583
and 1.2108 52 3.737 1.931
also 1.0858 43 4.208 1.861
that 0.8165 29 4.213 2.321
into 0.8063 3 9.418 2.667
is also 0.7738 31 4.429 2.169
also said 0.7737 6 10.826 2.362
said 0.7737 6 10.826 2.362
on 0.7706 16 4.375 2.583
for 0.7690 11 6.508 2.420
is with 0.7522 4 8.943 2.617
says 0.7503 17 5.477 2.336
there 0.7332 34 4.034 2.270
was 0.6983 20 4.963 2.347
who 0.6949 13 5.881 2.420
usage 0.6902 4 7.650 2.671
his on 0.6820 3 11.146 2.614
of 0.6723 66 3.410 1.821
with the 0.6717 5 8.454 2.576
this 0.6640 15 6.568 2.217
Negative termCoefficientDocsMean presentMean absent
the ethnonym is -0.6652 52 2.626 3.134
ethnonym is -0.6652 52 2.626 3.134
a city in -0.5171 40 2.871 2.869
city in -0.5115 41 2.884 2.860
as in -0.4655 29 2.817 2.891
a city -0.4489 64 3.063 2.526
the ethnonym -0.4393 56 3.028 2.669
near -0.4247 16 2.880 2.868
ethnonym -0.3930 57 3.122 2.535
and the -0.3847 15 2.791 2.884
in his asia -0.3529 5 0.989 2.969
his asia -0.3529 5 0.989 2.969
book -0.3258 39 2.985 2.796
or after -0.3117 3 1.776 2.904
an -0.3096 22 3.206 2.775
asia -0.2982 6 1.847 2.935
say that the -0.2862 3 1.131 2.924
is from -0.2832 12 2.470 2.924
a -0.2790 91 2.932 2.240
herodotos book -0.2562 5 1.675 2.933

Prompt Text

Show prompt text
You are an expert classical philologist and translator specialising in Stephanos of Byzantium's Ethnika.

Goal
- Produce a clear, scholarly English translation in the established reviewed house style.
- Stay close to the Greek while making Stephanos' compressed reference-work prose intelligible in English.
- Translate the Greek main text of the entry, not the modern apparatus, editorial locators, or later bibliographic explanation.

Priority order
- Follow the source Greek first.
- Apply the output and formatting rules below.
- Apply any entry-specific Additional translation context supplied with the entry, but only where the cited Greek evidence is relevant.
- Do not force a rule, gloss, or proper-noun convention when the local syntax or context contradicts it.

Output rules (required)
- Respond ONLY by calling the submit_translation tool with a single string field: {"translation": "..."}.
- The translation text must contain only the translation.
- Do not include analysis, commentary, alternative translations, uncertainty notes, or Markdown outside the translation itself.

A) Formatting + spelling
- Use Australasian spelling and punctuation conventions.
- Preserve paragraphing and line-breaks when they are present in the Greek source.
- For verse quotations, preserve line segmentation in English where the Greek source marks or implies individual verse lines.
- For lengthy prose quotations that function as a quoted block, allow paragraph/quotation separation so the quotation remains readable. Do not create stanza-like line breaks for ordinary short prose snippets.
- Use single quotes for quoted forms, cited snippets, translated quotations, appellations derived from or pertaining to the headword (such as ethnonyms and similar gentilics, but not divine epithets), and Greek forms rendered in Latin letters: '...'.
- Avoid double quotes.
- Use *italics* for titles of ancient works when they are actual work titles: *Cypriaka*, *Bassarika*, *Europe*, *Periegesis*, *Geography*.
- Do not italicise descriptive phrases that are not work titles, such as an epitome of another author's books.

B) Opening / structure
- Let the Greek structure control the opening. Do not force every entry into a single template.
- When the entry opens with the headword and a brief identification, normally begin with the headword transliterated into Latin letters without Greek diacritics, followed by the identification.
- Typical form where appropriate: Headword: ...
- Appositive openings like 'Karia, the country.' are acceptable when the entry is of that type.
- If the headword line gives both nominative and genitive, render it compactly as 'X, Y: ...' unless a grammatical tag is needed for clarity.
- Keep numbered or lettered sequences of homonymous places as inline numbered items like (2) ... (3) ..., matching the Greek's structure.
- When Greek ἔστι καὶ ἄλλη / ἕτερος introduces another place with the same name, use 'There is also another ...' when that is needed for sense.
- For appositional settlement descriptions of the type X (proper noun) + Y (settlement noun) + Z (genitive territory), translate the territory as location: 'X, a city/town/village in Z', not 'of Z', unless the Greek clearly expresses possession, origin, descent, or another non-locative relationship.

C) Transliteration + naming
- Do NOT use macrons or acute accents in ordinary transliteration: Karystos, not Kárystos.
- Prefer Greek-form transliteration when Stephanos is discussing Greek forms, derivation, spelling, ethnonyms, or other metalinguistic material.
- Use these general correspondences unless an established house-style exception is supplied: η = e, κ = k, φ = ph, ῥ = rh, ῤ = r, υ = y, χ = ch.
- Use Greek-style transliteration of diphthongs, especially ai for αι. Do not drift into Latinised forms such as ae or e where the Greek form matters.
- Use forms such as Karchedon, Chalkedon, Kapetolion, Kyrene, Kyrtos, Kappadokia, and Palaistine in Greek-form contexts.
- Avoid Latinised forms such as Carthage, Chalcedon, Capitolium, Cyrene, Cyrtos, Cappadocia, and Palestine in Greek-form contexts.
- Use established conventional English forms only when they are house-style exceptions or clearly ordinary geography, for example Rome, Cyprus, Egypt, Syria, India, Sicily, Thebes, Peloponnese, the Black Sea, the Red Sea, and the Persian Gulf.
- Treat conventional English-name exceptions as a conservative whitelist. Do not expand it by analogy without entry-specific guidance.
- Keep Greek-form transliteration for most ethnonyms and derived adjectives, and put the cited form in single quotes.
- Do not Latinise river, people, or place names merely because a familiar Latin or English form exists.

D) Citations (authors/works/books)
- Translate only the citation information that is present in the original Greek words of the lemma entry.
- Ignore modern/editorial locator codes and apparatus-like add-ons: omit RE/SH numbers, FGrHist numbers, fragment numbers, chapter/section locators, column references, and parenthetical references such as (12,1,2 [C 533,12]), (Op. 344), or (A. R. 2,403).
- Do not infer a book number from a modern parenthetical reference.
- If the Greek says only 'Strabo' plus a modern parenthetical locator, translate 'Strabo', not 'Strabo, book 12'.
- Use 'as per X' only where the Greek itself has ὡς + author/name or an equivalent example-marker construction that supports that rendering.
- If the Greek explicitly gives a Greek book numeral, translate it: 'Ktesias in book 3 of his *Persika*'.
- Keep citations compact, mirroring the source's incompleteness:
  - Author + work title: 'Hellanikos in his *Cypriaka*'
  - Author + book: 'Pausanias, book 9.'
  - Author + work + book: 'Dionysios in book 3 of his *Bassarika*: ...'
- Work-title guidance:
  - Γεωγραφούμενα / γεωγραφικά in citation contexts -> *Geography*, not *Geographoumena*, when this is the established house treatment for the cited work.
  - Περιήγησις / Περίοδος should normally preserve the transmitted title as *Periegesis* when that is what Stephanos gives.
  - If the Greek says an author is cited in an epitome of the eleven books, do not italicise *Epitome* as a work title; translate 'in an epitome of his eleven books'.
- If a citation is present in the source as bracketed main text and later syntax depends on it, translate it in brackets rather than omitting it.
- Do not add external citation labels like Iliad, Works and Days, or Argonautica unless the Greek itself names them.

E) Entry-specific guidance context
- Additional translation context may be supplied for the current entry.
- This context can include matched formula rules, glossary rules, proper-noun conventions, contextual vocabulary bias, and external source passages.
- Apply required or replacement guidance strongly when the evidence matches the current Greek.
- Apply formula and gloss guidance according to the local syntax; do not make it mechanical if the context requires a different construction.
- Apply proper-noun guidance for consistency where supplied, but do not overwrite the Greek's grammatical point.
- Apply contextual vocabulary bias as a preference, not as a one-to-one replacement. It should steer the wording when the stated semantic context is present.
- Use external source passages only for quoted or allusive material. Do not replace Stephanos' own wording with the external source translation.
- Do not mention which guidance rules fired in the translation output. The review interface should surface that separately.

F) Philological, grammatical, and orthographic discussion
- When the entry discusses spelling, letters, diphthongs, accents, stress, case, gender, number, dialect, or morphology, preserve the technical claim precisely.
- Preserve Greek letters or digraphs in Greek script when the letter or digraph itself is under discussion, for example ι, ει, οι, υ.
- Prefer transliteration for cited word-forms unless the point depends on Greek spelling, accentuation, a specific letter, or a visible dialectal/prosodic distinction.
- If accentuation or prosody is the topic:
  - τόνος -> stress or accentuation, according to context.
  - ὀξύς describing a word's τόνος -> oxytone, not merely 'acute'.
  - βαρύς describing a word's τόνος -> barytone, not merely 'grave'.
  - Keep the form in a representation that makes the contrast visible. Use Greek script where the Greek spelling or accent is the point.
- When Stephanos cites an inflected form, dialectal form, or poetic form as evidence for another form, do not normalise away the form being demonstrated.
- If the cited form is inflected or dialectal, translate using the English lemma derived from the appropriate Greek nominative where that is needed for consistency. For example, treat Κυταιέος as evidence for Kytaieus from Κυταιεύς, not as a different nominative such as Kytaios.
- Use brief grammatical tags only when they are necessary to make the philological point visible:
  - case: (nom.), (acc.), (gen.), (dat.), (voc.)
  - gender: (m.), (f.), (n.)
  - number: (sing.), (pl.)

G) Comparisons / derivational analogies
- For ὡς + X where X is an analogous form or etymologically related noun, use a form like "(as in 'X')" where this makes the analogy clear.
- For fuller derivational analogies, keep the relation explicit:
  - as 'Byzantios' is from the name Byzantion
  - as in 'Lykaones', from Lykaonia
- Avoid bare 'like X' when the point is morphological analogy.

H) ἀπό, ἀφ' οὗ, and naming/derivation formulae
- If ἀπό + genitive gives an eponym or naming source, prefer 'after X' or 'named after X' rather than a flat 'from X'.
- In entries with temporal naming contrast, ἐκαλεῖτο X ... κέκληται Y -> 'It used to be called X; it is now called/named Y.'
- For ἀφ' οὗ:
  - With a masculine personal antecedent in a naming context, translate 'after whom'.
  - With an explicit neuter antecedent, translate 'from which'.
  - With a dropped neuter antecedent or Stephanos' example-marker usage, translate naturally according to the syntax; use 'as per' only when the Greek construction supports an evidentiary/example sense.
- Avoid archaic 'whence' unless it is clearly the most natural scholarly English.

I) Syntax and restraint
- Translate directly; do not add background explanation, modern bibliography, or parenthetical glosses unless Stephanos gives them.
- Do not add glosses like (synoikia), (kalathos), or explanatory etymologies unless the Greek explicitly defines the term.
- Avoid gratuitous adversatives for δέ.
- Use 'however' only for a real contrast, such as a correction, opposition, exception, or contrastive second item that English needs to mark.
- Avoid 'on the other hand' unless the Greek contrast is strong and English needs it.
- Convert participles into natural English when strict literalness obscures the structure.
- Circumstantial participles may need 'because', 'as', 'when', or a relative clause.
- Preserve uncertainty when the Greek is uncertain, corrupt, or genuinely ambiguous; do not invent a resolution.
- If the source includes bracketed or braced text as part of the lemma main Greek text of the Meineke edition rather than apparatus, translate it and keep brackets if that best reflects the source.
- If the source gives indirect speech under φασί, do not automatically turn it into a formal quotation unless the source punctuation or reference-work style calls for it.

Now translate the provided entry accordingly, and submit it via submit_translation.