paper_guidance_ablation_gpt56 v3 Evaluation

Generated: 2026-09-20 11:23:06 UTC

Pairs100
Distinct lemmas100
Slope0.947
R^20.989
BLEU-456.1%
chrF++75.7%
METEOR81.7%
ROUGE-L81.0%
BERTScore96.9%
COMET81.2%
BLEURT78.8%

Profile version id: 2588

First translation using this version: 2026-07-18

Total usable AI runs for this version: 300 runs across 100 lemmas

Created: 2026-07-18 21:14:55.545407+10:00

Active: True

Notes: Arm C: Stephanos v3 static prompt shell plus matched entry-specific guidance. Model gpt-5.6-sol; Responses API; medium reasoning; top_p 1.0; three runs per Kappa-corpus entry.

Prompt length scatter plot

Length Regression

FormulaAI words = intercept + slope * human words
Slope0.947
Intercept1.738
R^20.989
P-value1.15e-97
Pearson r0.994
Mean source words30.400
Length fallback rows0
Mean human words44.150
Mean AI words43.550
Mean absolute length residual2.712
Mean absolute percent error6.4%

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 = -0.440, R^2 = 0.193, p = 4.76e-06.
chrF++ vs passage length
chrF++: significant negative; r = -0.474, R^2 = 0.225, p = 6.19e-07.
METEOR vs passage length
METEOR: significant negative; r = -0.488, R^2 = 0.239, p = 2.53e-07.
ROUGE-L vs passage length
ROUGE-L: significant negative; r = -0.457, R^2 = 0.209, p = 1.71e-06.
BERTScore vs passage length
BERTScore: significant negative; r = -0.633, R^2 = 0.400, p = 1.67e-12.
COMET vs passage length
COMET: significant negative; r = -0.457, R^2 = 0.209, p = 1.74e-06.
BLEURT vs passage length
BLEURT: significant negative; r = -0.705, R^2 = 0.496, p = 2.82e-16.
Trigram precision vs passage length
Trigram precision: significant negative; r = -0.350, R^2 = 0.123, p = 0.0004.
Trigram recall vs passage length
Trigram recall: significant negative; r = -0.380, R^2 = 0.144, p = 9.83e-05.
Trigram F1 vs passage length
Trigram F1: significant negative; r = -0.365, R^2 = 0.133, p = 0.0002.
Trigram Jaccard vs passage length
Trigram Jaccard: significant negative; r = -0.375, R^2 = 0.140, p = 0.0001.
Metric Rows Pattern Pearson r R^2 P-value Slope Status
BLEU-4 100 significant negative -0.440 0.193 4.76e-06 -0.00289 ok
chrF++ 100 significant negative -0.474 0.225 6.19e-07 -0.00192 ok
METEOR 100 significant negative -0.488 0.239 2.53e-07 -0.00202 ok
ROUGE-L 100 significant negative -0.457 0.209 1.71e-06 -0.00183 ok
BERTScore 100 significant negative -0.633 0.400 1.67e-12 -0.00042 ok
COMET 100 significant negative -0.457 0.209 1.74e-06 -0.00080 ok
BLEURT 100 significant negative -0.705 0.496 2.82e-16 -0.00234 ok
Trigram precision 100 significant negative -0.350 0.123 0.0004 -0.00279 ok
Trigram recall 100 significant negative -0.380 0.144 9.83e-05 -0.00296 ok
Trigram F1 100 significant negative -0.365 0.133 0.0002 -0.00287 ok
Trigram Jaccard 100 significant negative -0.375 0.140 0.0001 -0.00320 ok

Translation Similarity Metrics

Mean BLEU-456.1%
Mean chrF++75.7%
Mean METEOR81.7%
Mean ROUGE-L81.0%
Mean BERTScore96.9%
Mean COMET81.2%
Mean BLEURT78.8%
Legacy smoothed corpus BLEU54.4%
Unigram F181.3%
Bigram F160.2%
Trigram precision47.9%
Trigram recall47.3%
Trigram F147.6%
Trigram Jaccard31.2%
4-gram F138.5%

Largest Length Residuals

Lemma ID Headword Human words AI words Residual BLEU-4 chrF++ METEOR ROUGE-L BERTScore COMET BLEURT Human Run Model
2604 Καρχηδών 117 128 15.457 30.5% 61.6% 64.3% 63.2% 93.2% 77.6% 64.9% reviewed/approved 14901 gpt-5.6-sol
2484 Καρία 270 246 -11.442 45.8% 67.5% 67.1% 72.5% 93.0% 81.7% 55.8% reviewed/approved 14856 gpt-5.6-sol
2609 Κάσπειρος 135 119 -10.590 42.7% 65.7% 68.2% 72.4% 91.8% 73.4% 68.6% reviewed/approved 14928 gpt-5.6-sol
2603 Κάρυστος 175 178 10.528 42.1% 66.2% 75.4% 71.8% 94.7% 78.5% 63.0% reviewed/approved 14892 gpt-5.6-sol
2055 Καβασσός 109 95 -9.966 39.6% 61.8% 62.9% 65.7% 95.2% 77.4% 54.1% reviewed/approved 14424 gpt-5.6-sol
7251 Κύρτος 66 55 -9.243 40.1% 62.7% 69.6% 72.7% 95.6% 73.9% 70.1% reviewed/approved 15162 gpt-5.6-sol
2342 Κάμιρος 54 62 9.122 45.2% 73.0% 79.3% 77.6% 95.7% 80.2% 73.9% reviewed/approved 14784 gpt-5.6-sol
2328 Καλλίπολις 43 50 7.539 52.7% 71.5% 78.0% 75.3% 95.2% 77.1% 71.7% reviewed/approved 14712 gpt-5.6-sol
2470 Καππαδοκία 68 73 6.863 26.0% 60.3% 85.9% 82.3% 95.2% 74.7% 72.1% reviewed/approved 14838 gpt-5.6-sol
2119 Κάλλατις 65 70 6.704 31.3% 61.2% 63.4% 56.3% 94.4% 71.6% 66.6% reviewed/approved 14676 gpt-5.6-sol
2455 Κάναστρον 60 52 -6.561 31.5% 65.2% 63.6% 71.4% 95.0% 72.8% 65.6% reviewed/approved 14811 gpt-5.6-sol
7259 Κύφος 54 59 6.122 52.5% 73.7% 89.2% 81.4% 95.8% 76.1% 76.9% reviewed/approved 15234 gpt-5.6-sol
7254 Κύτα 84 87 5.710 53.4% 75.4% 81.8% 79.5% 95.9% 80.1% 76.4% reviewed/approved 15189 gpt-5.6-sol
2116 Κάληρος 40 34 -5.620 19.4% 52.6% 59.0% 62.2% 93.7% 76.0% 70.3% reviewed/approved 14649 gpt-5.6-sol
7255 Κυτέριον 22 28 5.427 48.4% 71.2% 67.6% 68.0% 96.1% 76.4% 75.5% reviewed/approved 15198 gpt-5.6-sol
2626 Κατάνη 95 97 5.292 39.2% 62.4% 65.8% 65.6% 94.2% 74.7% 65.2% reviewed/approved 15000 gpt-5.6-sol
2605 Κάσιον 56 60 5.227 55.6% 76.1% 85.6% 77.6% 96.7% 78.6% 74.8% reviewed/approved 14910 gpt-5.6-sol
7266 Κώμη 79 81 4.445 25.6% 52.9% 49.8% 55.0% 93.8% 75.4% 58.2% reviewed/approved 15297 gpt-5.6-sol
3530 Κριώα 26 22 -4.361 38.8% 68.3% 64.3% 75.0% 96.4% 77.9% 77.5% reviewed/approved 15072 gpt-5.6-sol
7267 Κωνώπη 64 58 -4.349 45.3% 71.3% 70.1% 78.7% 97.1% 79.2% 79.2% reviewed/approved 15306 gpt-5.6-sol
7247 Κύρνος 50 45 -4.090 32.3% 56.3% 63.6% 61.1% 95.6% 79.1% 77.8% reviewed/approved 15126 gpt-5.6-sol
7258 Κύτωρος 30 34 3.851 66.1% 87.0% 96.3% 87.5% 97.8% 82.6% 82.6% reviewed/approved 15225 gpt-5.6-sol
2621 Κάστνιον 25 29 3.586 45.3% 75.7% 72.9% 74.1% 96.8% 81.6% 78.6% reviewed/approved 14946 gpt-5.6-sol
2477 Καρδαμύλη 43 46 3.539 47.0% 78.2% 88.4% 78.7% 96.6% 82.5% 79.6% reviewed/approved 14847 gpt-5.6-sol
7260 Κυχρεῖος πάγος 61 63 3.492 52.4% 71.3% 82.6% 72.6% 96.4% 78.1% 70.6% reviewed/approved 15243 gpt-5.6-sol
2115 Καλὴ ἀκτή 64 59 -3.349 36.3% 63.4% 73.2% 71.5% 95.6% 76.9% 68.5% reviewed/approved 14640 gpt-5.6-sol
3288 Κορώνεια 110 109 3.087 39.3% 67.7% 69.7% 72.7% 95.7% 78.3% 69.2% reviewed/approved 15045 gpt-5.6-sol
7245 Κύρης 13 17 2.951 50.3% 77.6% 87.9% 80.0% 96.4% 76.6% 79.3% reviewed/approved 15108 gpt-5.6-sol
2075 Καινίνη 15 13 -2.943 65.3% 80.4% 80.4% 85.7% 98.2% 85.9% 79.7% reviewed/approved 14514 gpt-5.6-sol
7261 Κύψελα 34 31 -2.937 75.8% 83.1% 91.8% 89.2% 98.9% 83.5% 83.7% reviewed/approved 15252 gpt-5.6-sol

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: 17.013; cross-validated R^2: -0.101.

Positive termCoefficientDocsMean presentMean absent
απο 0.4406 27 2.217 -0.820
και 0.4321 66 0.075 -0.146
δε και 0.3842 8 2.894 -0.252
ος 0.3642 6 5.352 -0.342
αφ 0.3590 6 4.528 -0.289
παιδος 0.3169 4 4.908 -0.205
πολιτης 0.3165 18 2.625 -0.576
αφ ου 0.3121 5 4.969 -0.262
εστι δε και 0.3114 3 7.361 -0.228
εν τη 0.3080 7 3.526 -0.265
λεγεται 0.3032 9 3.286 -0.325
εις 0.2965 7 2.528 -0.190
λεγεται και 0.2694 8 3.255 -0.283
τη 0.2492 15 1.215 -0.214
ου και 0.2477 4 6.263 -0.261
εκαλειτο δε 0.2463 3 4.848 -0.150
εστι δε 0.2439 5 4.185 -0.220
και αλλη 0.2396 3 8.820 -0.273
της 0.2133 35 0.453 -0.244
εκληθη 0.2064 4 5.433 -0.226
Negative termCoefficientDocsMean presentMean absent
τα -0.3382 12 -2.318 0.316
το -0.2885 70 -0.301 0.703
διονυσιος -0.2862 4 -5.343 0.223
τοπικα -0.2538 3 -5.278 0.163
τα τοπικα -0.2538 3 -5.278 0.163
θρακης -0.2477 5 -2.812 0.148
καππαδοκιας -0.2186 3 -3.702 0.114
ως το -0.2067 3 -5.672 0.175
εστι και -0.2048 20 -0.347 0.087
θεοπομπος -0.1969 5 -1.628 0.086
ουτως -0.1954 5 -4.340 0.228
γαρ -0.1932 9 -1.622 0.160
το εθνικον -0.1901 56 -0.292 0.372
εθνικον -0.1901 56 -0.292 0.372
μεταξυ -0.1765 5 -2.194 0.115
και του -0.1734 4 -3.446 0.144
λυκιας -0.1706 3 -4.073 0.126
του δε -0.1669 3 -4.090 0.127
τινες -0.1663 6 -1.794 0.115
τε -0.1639 5 1.159 -0.061

Greek source terms predicting large absolute residuals

Features: 176; ridge alpha: 2.031; cross-validated R^2: -0.044.

Positive termCoefficientDocsMean presentMean absent
δε 2.3241 46 3.849 1.744
εκαλειτο 2.0655 7 7.840 2.326
εκαλειτο δε 1.7026 3 12.476 2.410
τα 1.6621 12 6.088 2.252
απο 1.5546 27 4.615 2.008
και 1.4933 66 3.290 1.591
τε 1.4826 5 5.683 2.556
δε και 1.4111 8 6.423 2.389
τω 1.3128 16 4.318 2.406
απο της 1.2631 6 7.468 2.408
τη 1.2535 15 4.875 2.330
υπο 1.1542 4 5.818 2.583
εκαλειτο δε και 1.1023 3 12.476 2.410
αλλ 1.0883 3 8.004 2.548
εστι δε και 1.0296 3 7.361 2.568
ος 0.9906 6 5.678 2.523
γαρ 0.9554 9 6.567 2.331
της 0.9391 35 4.031 2.002
εν 0.9332 37 3.470 2.267
μεν 0.9281 4 7.250 2.523
Negative termCoefficientDocsMean presentMean absent
νησος -1.1140 9 1.070 2.874
και το -1.0548 9 2.667 2.716
στραβων -1.0081 15 2.466 2.755
το εθνικον -0.9055 56 2.470 3.020
εθνικον -0.9055 56 2.470 3.020
μια -0.8812 3 0.432 2.782
προς τη -0.7914 4 1.457 2.764
οικητωρ -0.6975 6 3.172 2.683
προς -0.6585 15 2.786 2.699
και νησος -0.6324 3 0.384 2.784
και πολις -0.6050 5 2.144 2.742
το -0.5849 70 2.668 2.815
παυσανιας -0.5670 4 1.324 2.770
αυτην -0.5649 9 2.615 2.722
ποσειδωνος -0.5642 3 1.766 2.741
εθνος -0.5620 8 2.995 2.687
θαλασση -0.5310 3 0.842 2.770
τινες -0.5209 6 1.794 2.771
ακρα -0.5082 4 1.180 2.776
θηλυκον και -0.5056 3 1.505 2.749

AI English terms predicting large absolute residuals

Features: 332; ridge alpha: 0.702; cross-validated R^2: 0.120.

Positive termCoefficientDocsMean presentMean absent
the 3.9638 93 2.814 1.358
he 3.5697 4 10.089 2.405
was 3.4562 19 6.040 1.931
their 2.5202 4 10.231 2.399
them 2.4910 7 6.848 2.401
renowned 2.2705 4 9.779 2.418
and 2.2036 56 3.655 1.512
to 2.1491 21 4.875 2.137
among 2.1239 6 5.985 2.503
kappadokia 2.1156 4 5.259 2.606
says 1.9969 14 4.891 2.357
for 1.9908 13 5.170 2.345
used 1.9761 8 7.190 2.323
his on 1.9472 3 9.408 2.505
was called 1.8865 6 9.418 2.284
was named 1.8349 7 7.594 2.345
when 1.7278 7 7.461 2.355
in his on 1.6294 3 9.408 2.505
however 1.5716 7 5.656 2.490
homer 1.5205 3 7.542 2.563
Negative termCoefficientDocsMean presentMean absent
island -1.6665 14 1.541 2.903
it is -1.5949 18 2.523 2.753
as in -1.5903 26 2.642 2.737
is called -1.4840 7 1.278 2.820
one of the -1.4607 3 0.432 2.782
calls -1.4279 4 0.887 2.788
one of -1.4151 5 0.729 2.816
strabo -1.3246 15 2.466 2.755
it -1.2792 41 3.729 2.005
in libya -1.1464 4 0.920 2.787
one -1.1380 11 2.192 2.776
the ethnonym -1.1171 52 2.588 2.846
strabo book -0.9793 9 1.748 2.807
calls it -0.9608 3 1.091 2.762
poseidon -0.9458 3 1.766 2.741
an island -0.9146 10 1.099 2.891
ethnonym is -0.9056 52 2.604 2.829
the ethnonym is -0.8957 51 2.612 2.816
near -0.8545 19 2.460 2.771
is from the -0.8484 5 1.496 2.776

Human English terms predicting large absolute residuals

Features: 335; ridge alpha: 2.031; cross-validated R^2: 0.014.

Positive termCoefficientDocsMean presentMean absent
the 2.5025 90 2.866 1.328
and 1.8289 52 3.664 1.680
was 1.5802 20 5.100 2.115
of 1.4241 66 3.227 1.713
who 1.3667 13 5.414 2.308
this 1.1900 15 5.940 2.142
on 1.1849 16 4.646 2.344
with the 1.1456 5 8.562 2.404
used to be 1.0839 6 8.955 2.313
his on 0.9988 3 9.408 2.505
was called 0.9730 3 10.107 2.483
also said 0.9605 6 8.018 2.373
said 0.9605 6 8.018 2.373
as 0.9412 60 3.109 2.117
their 0.9120 3 12.476 2.410
to be called 0.8705 5 8.641 2.400
be 0.8693 15 5.428 2.233
to 0.8585 31 3.757 2.243
after 0.8523 22 4.886 2.099
in his on 0.8495 3 9.408 2.505
Negative termCoefficientDocsMean presentMean absent
as in -0.8960 29 2.408 2.836
strabo -0.7809 15 2.466 2.755
island -0.7792 14 1.541 2.903
the ethnonym is -0.7483 52 2.474 2.969
ethnonym is -0.7483 52 2.474 2.969
those -0.6932 5 1.174 2.793
one of -0.6426 5 1.746 2.763
it is -0.6370 24 2.976 2.629
the ethnonym -0.6361 56 2.623 2.825
one of the -0.6280 4 1.647 2.756
an -0.5732 22 2.201 2.856
and the -0.5648 15 2.955 2.669
an island -0.5275 9 1.200 2.862
ethnonym -0.5250 57 2.752 2.659
or after -0.5142 3 2.262 2.726
in the -0.5090 25 3.171 2.559
it is also -0.5025 3 0.372 2.784
people -0.4758 9 2.809 2.702
is from the -0.4596 3 1.546 2.748
book -0.4526 39 2.324 2.960

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.