Asymmetric Quantization: Near-Lossless Retrieval with 97% Storage Reductionmixedbread.com·110 pts·breadislove·44
The whole point of OpenAI's Responses API is to help them hide reasoning tracesseangoedecke.com·2 pts·breadislove·0
Boosting Claude: Faster, Clearer Code Analysis with MGrepelite-ai-assisted-coding.dev·1 pts·breadislove·0
Show HN: Semantic search over the National Gallery of Artnga.demo.mixedbread.com·145 pts·breadislove·37
Unlock DeepSeek-R1 Inference Performance on AMD Instinct MI300X GPUrocm.blogs.amd.com·2 pts·breadislove·0
An Analysis of Chinese LLM Censorship and Bias with Qwen 2 Instructhuggingface.co·3 pts·breadislove·0
Most AI startups are service businesses, pretending to be product companiesnextword.substack.com·7 pts·breadislove·0