Google Finds ‘Refusal to Translate’ Most Common Form of LLM Verbosity
Google researchers reveal that verbose large language models, which offer multiple translations or refuse to...
Array
(
[post_type] => post
[posts_per_page] => 12
[paged] => 3
[post_status] => publish
[tax_query] => Array
(
[relation] => AND
[0] => Array
(
[taxonomy] => post_tag
[field] => term_id
[terms] => 525
[operator] => IN
)
)
)
Google Aligns LLM Translation with Human Translation Processes
Google’s new approach mimics the human translation process, incorporating research, drafting, and iterative refinement in...
Google Translate Almost Doubles its Language Coverage Overnight
Google leveraged its large language model, PALM2, to roll out 110 new languages, extending its...
Google Shares Current View on Using AI for Website Translation
Google SEO specialist provides an updated take on how web pages are ranked if creators...
Google Presents the First On-Device, Real-Time Speech-to-Speech Translation Model
Researchers from Google present the first model for real-time speech-to-speech translation on mobile devices, surpassing...
10 Large Language Models That Matter to the Language Industry
An overview of ten recent large language models, with breakthrough translation, voice AI, and speech-to-speech...
Is Translation Quality Evaluation a Solved Problem?
Automated translation evaluation saves time, but eliminating the need for a final human quality check...
OpenAI Cautiously Debuts Multilingual Generative AI Voice Engine
From the creators of ChatGPT, Silicon Valley’s OpenAI introduces Voice Engine, whose abilities include multilingual...
Google Translate No Longer Needs Humans to Improve Translations
The world’s leading machine translation tool no longer accepts human input through its Contribute feature...
Alibaba Echoes Google’s Call to Move Away From Pre-Translation
Researchers from Alibaba and Singapore’s NTU argue that direct inference is more effective for culture-related...
How Google Wants to Transcend Pre-Translation in Large Language Models
Researchers from Google suggest that embracing direct inference over pre-translation enhances large language model performance...