Alibaba Qwen3.5 LiveTranslate: 2.8s Latency Breakthrough
Alibaba's Qwen3.5-LiveTranslate-Flash slashes real-time translation latency to 2.8 seconds while preserving speaker voic…
🏷️ llm
22 articles about 'llm'
δ-mem: Cutting LLM Memory Costs by 90%
New δ-mem framework slashes GPU memory usage for LLMs by 90%, enabling efficient online inference on consumer hardware.
Amazon Lex Assisted NLU Boosts Bot Accuracy
AWS launches Assisted NLU for Amazon Lex, leveraging LLMs to improve intent recognition and slot filling without manual …
DeepSeek Chat Glitch Exposes V2K Forum Content
DeepSeek's AI model accidentally outputs explicit content from China's V2EX forum, raising data privacy and training set…
Brix Seeks AI Engineers for Agentic Recruiting OS
Brix is hiring AI engineers to build autonomous recruiting agents. This role focuses on LLM reasoning and multi-agent wo…
Dev Challenges LLM Embeddings: 72-Hour Experiment
A developer systematically refutes three hypotheses on semantic units using geometric algebra and factor attention on du…
Does Correcting AI Update Its Knowledge?
Correcting an AI in chat does not instantly update its model. Learn how training data cycles and RAG systems impact long…
Xiaomi MiMo Orbit Hits 80T Tokens
Xiaomi's MiMo Orbit initiative distributes nearly 80 trillion tokens in under two weeks, signaling aggressive expansion …
Democratizing AI: Fine-Tuning LLMs on Consumer GPUs
Developers now fine-tune powerful small language models on consumer hardware, reducing costs and boosting privacy for lo…
AI Judges Rank Show HN Posts via TrueSkill
New study uses LLM judges and TrueSkill to rank 1,000 Show HN posts by merit.
DeepSeek R1: Open-Source AI Challenges OpenAI
DeepSeek releases R1 model, offering open-source reasoning capabilities that rival top proprietary models at a fraction …
Panasonic Brings LLMs to Industrial IoT Edge
Panasonic integrates large language models into its industrial IoT edge devices, enabling real-time AI inference on fact…