LLMs Show Variable Reliance on Supplied Lexical Knowledge in Translation Tasks
Study finds seven of nine models favor supplied meanings over prior knowledge even when instructed to ignore them.
Stories per day · 30 days
Relatedgranite2ibm2game-engine1godot1lexical-knowledge1open-weights1out-of-bounds1speech1speech-recognition1transcription1
6 stories, newest first · one card per story
Stories tagged #translation
Godot 4.x translation lookup heap over-read fix merged
Upstream pull request 124701 checks bucket and string offsets used by message lookup and the translated message list.
IBM releases Granite Speech 5.0 2B for transcription and translation
The captured page is a cookie notice, so size, languages, license, and availability are not confirmed beyond the title.
IBM announces Granite Speech 5.0 for speech tasks
The Hugging Face post title says the model transcribes, translates, and attributes speech more accurately and faster.
AI for everyone in every language
Google says its AI now spans 300+ languages reaching 7 billion people, unveiling Gemini 3.5 Transcribe, Live Translate, and TranslateGemma models.
Drift-Constrained Optimization: Only Direction Matters in Fine-Tuning Instruct Models
Drift-Constrained Optimization reformulates fine-tuning as update-direction selection, letting Qwen3 models improve target tasks within a behavioral drift budget.