Unverified50% confidenceFactExact time
Vocabulary surgery reduces the merged vocabulary to a target of 100k, saving about 649 MiB, at the cost of making the model strictly English monolingual.
1
Sources
50%
Confidence
Long-term
Relevance
10/30/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedWhen 'write in STE style' is specified, the model automatically follows the constraints of Simplified Technical English (about 900 core words, sentences usually no more than 20 words, one meaning per word), producing highly consistent, unambiguous text.67% similarUnverifiedالهدف الجوهري من خطوات التطبيع هو تقليل تشتت المفردات (vocabulary sparsity) أي تقليص عدد الصيغ المختلفة للكلمة نفسها65% similarUnverifiedRico Sennrich et al. quantified in their 2016 paper "Neural Machine Translation of Rare Words with Subword Units": on a 100-million-word German corpus, keeping whole words requires a vocabulary of 1.75 million entries, and the test set still leaves 1,079 unknown words; splitting into characters requires about 3,000 symbols, but the corpus balloons to 550 million positions; the subword method BPE uses about 63,000 symbols, 112 million positions, and zero unknown words.63% similarUnverifiedWordCraft is currently in alpha; the author estimates that by command count it covers 88% of Word's ribbon features, but it is only about 60% of the way to truly replacing Word, and for everyday basic needs that figure is 63%.61% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/1115183API
curl https://kongchang.com/api/v1/knowledge/claims/1115183MCP
get_claim(id=1115183)