For the last paper read of the year, Arize CPO & Co-Founder, Aparna Dhinakaran, is joined by a Dat Ngo (ML Solutions Architect) and Aman Khan (Product Manager) for an exploration of the new kids on the block: Gemini and Mixtral-8x7B.
There's a lot to cover, so this week's paper read is Part I in a series about Mixtral and Gemini. In Part I, we provide some background and context for Mixtral 8x7B from Mistral AI, a high-quality sparse mixture of experts model (SMoE) that outperforms Llama 2 70B on most benchmarks with 6x faster inference Mixtral also matches or outperforms GPT3.5 on most benchmarks. This open-source model was optimized through supervised fine-tuning and direct preference optimization.
Stay tuned for Part II in January, where we'll build on this conversation in and discuss Gemini-developed by teams at DeepMind and Google Research.
Link to transcript and live recording: https://arize.com/blog/a-deep-dive-into-generatives-newest-models-mistral-mixtral-8x7b/
To learn more about ML observability, join the Arize AI Slack community or get the latest on our LinkedIn and Twitter.
Breaking Down EvalGen: Who Validates the Validators?
Keys To Understanding ReAct: Synergizing Reasoning and Acting in Language Models
Demystifying Chronos: Learning the Language of Time Series
Anthropic Claude 3
Reinforcement Learning in the Era of LLMs
Sora: OpenAI’s Text-to-Video Generation Model
RAG vs Fine-Tuning
Phi-2 Model
HyDE: Precise Zero-Shot Dense Retrieval without Relevance Labels
How to Prompt LLMs for Text-to-SQL: A Study in Zero-shot, Single-domain, and Cross-domain Settings
The Geometry of Truth: Emergent Linear Structure in LLM Representation of True/False Datasets
Towards Monosemanticity: Decomposing Language Models With Dictionary Learning
RankVicuna: Zero-Shot Listwise Document Reranking with Open-Source Large Language Models
Explaining Grokking Through Circuit Efficiency
Large Content And Behavior Models To Understand, Simulate, And Optimize Content And Behavior
Skeleton of Thought: LLMs Can Do Parallel Decoding
Llama 2: Open Foundation and Fine-Tuned Chat Models
Lost in the Middle: How Language Models Use Long Contexts
Orca: Progressive Learning from Complex Explanation Traces of GPT-4
Create your
podcast in
minutes
It is Free
The Universe Speaks in Numbers
Breaking Math Podcast
Opinionated History of Mathematics
Biostatistics Podcast
SOA Podcasts - Society of Actuaries