After 500M tokens of Polish-heavy continued pretraining with its original tokenizer, Qwen2.5-1.5B's Polish likelihood multiple-choice accuracy rises on neither translation-paired Belebele items (difference -0.1171, 95% interval -0.1605 to -0.0702) nor translation-paired MMLU items (difference 0.0000, 95% interval -0.0458 to 0.0424).

Relation: Raised on a count of benchmarks

metric

Likelihood multiple-choice accuracyQuantity compared.

language

PolishLanguage of the items.

subject

Qwen2.5-1.5B continued with its own tokenizerModel after continued pretraining with its original tokenizer.

comparator

Qwen2.5-1.5BModel before continued pretraining.

count raised

0 benchmarksBenchmarks where the subject's Polish accuracy is higher with an interval above zero.

count total

2 benchmarksBenchmarks compared.

benchmarks

translation-paired Belebele items; translation-paired MMLU itemsBenchmarks compared.

training

500,170,752 Qwen tokens of continued pretraining of each arm on the same document sequence, constant learning rate before any decay anneal, one run per armTraining the continued-pretraining arms received.

Method and evaluation protocol

Count the benchmarks whose paired bootstrap 95% interval of the Polish accuracy difference lies above zero.

Dataset

Translation-paired Belebele and MMLU items after the exclusions.Version: unspecified · Access: public

Reported results

0 of 2 benchmarks.

Uncertainty and replication

Paired bootstrap 95% intervals per benchmark.

Evidence references

analysis.json · publichttps://github.com/stw2/tokenizer-science-tax/blob/5e8117ffb3f168331ee63f1032cd40a62c6296bd/experiments/E11-cross-language-access/results/analysis.jsonmetrics.json · publichttps://github.com/stw2/tokenizer-science-tax/blob/5e8117ffb3f168331ee63f1032cd40a62c6296bd/experiments/E11-cross-language-access/results/metrics.json

Limitations

One model family at 1.5B parameters; the checkpoints are pre-anneal, from one training run each. English MMLU items may be in Qwen2.5 pretraining data, so MMLU gaps can be inflated; Polish MMLU items are machine translations. The 13 excluded pairs were chosen after the base model's cells had been scored. Continued pretraining was 500M tokens without annealing; longer training may differ. Letter priors: on Polish MMLU items the checkpoint that kept the original tokenizer predicts A for 4 of 600 items and the APT4 checkpoint predicts B for 437, against 134 to 160 gold items per letter; on Polish Belebele items the APT4 checkpoint predicts B for 360 and the other checkpoint D for 326 of 600; the base model's predictions are the least concentrated (all scored rows, results/metrics.json pred_count_*).