TheSequence - Gemma 2: A Release That Matters
Was this email forwarded to you? Sign up here Next Week in The Sequence:
You can subscribe to The Sequence below:📝 Editorial: Gemma 2: A Release That MattersSmall language models (SLMs) are gaining a lot of momentum in both research and product releases. SLMs have the potential to streamline the adoption of generative AI and unlock use cases such as on-device AI. Most of the SLM releases from the past year have been limited to the models themselves, and that’s hardly enough. Adopting SLMs (just like larger models) comes with significant challenges. Interpretability and security certainly rank at the top of the list. Last week, Google took a major step in addressing some of these challenges with a new release of its Gemma stack. In June, Google announced the release of Gemma 2, a series of mid-size models with 9 billion and 27 billion parameters, respectively. A few days ago, Google extended that release with some exciting additions:
The Gemma release is significant because it is open source. This should encourage researchers and developers to expand on these capabilities, particularly in the case of Gemma Scope, given that interpretability is such a challenge in generative AI applications. I have been tinkering with Gemma Scope over the past few days, and it’s pretty useful. With this release, it seems that Gemma is evolving from a family of models to complete stacks to enable SLM adoption in real-world applications. This release definitely matters. What’s the best model for RAG? Our latest LLM Hallucination Index ranks 22 of the leading models on their performance across 3 different RAG tasks, evaluating the correctness of their responses and propensity to hallucinate. See which model comes out on top and why larger is not always better…" 🔎 ML ResearchGemma 2Google DeepMind published a paper detailing Gemma 2, a new set of small models ranging from 2 to 27 billion parameters. Gemma 2 extends its predecessors with new techniques such as interleaving local-global attentions and group query attention —> Read more. SAM 2Meta AI published a paper introducing the second version of its Segment Anything Model(SAM) for real time object segmentation. SAM expands over its predecesors by providing an unified model for object segmentation in videos and images —> Read more. TraceMicrosoft Research published a paper and open source code of Trace, a framework for AI systems optimization. Trace is a new AutoDiff like tool but can be used in systems without gradients —> Read more. CMU-MATHCarnegie Mellon University researchers published details about CMU-MATH, a model that took second place in the AI mathematical olympiad. CMU-MATH uses a dual model system that which includes a policy model that produces multiple solutions and a reward model that chooses the answer with the highest weight —> Read more. MoMaMeta AI publishd a paper introducing MoMa, a mixture-of-experts architecture for mixed model models. MoMa is based on the Chameleon architecture and processes images and text in arbitrary sequences —> Read more. Berkeley HumanoidAI researchers from UC Berkeley published a paper introducing Berkeley Humanoid, a research framework for learning based control. The framework includes a robot designed to learn algorithm with low simulation complexity —> Read more. 🤖 AI Tech ReleasesGemma 2Google open sourced new additions to its Gemma models including a 2B parameter model, a guardrail framework and an interpretability tool —> Read more. torchchatThe PyTorch team released torchchat, a library for accelerated inference in laptop, mobile and desktop devices —> Read more. Stable Fast 3DStability AI open sourced Stable Fast 3D, a model for rapid 3D asset generation —> Read more. 🛠 Real World AIAgents in the EnterpriseSalesforce shares some perspectives about the impact of AI agents in enterprise automation tasks —> Read more. 📡AI Radar
You’re on the free list for TheSequence Scope and TheSequence Chat. For the full experience, become a paying subscriber to TheSequence Edge. Trusted by thousands of subscribers from the leading AI labs and universities. |
Older messages
📽 [Webinar] Beat GPT-4 with a Small Model and 10 Rows of Data*
Friday, August 2, 2024
Small language models (SLMs) are increasingly rivaling the performance of large foundation models like GPT-4. ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
Edge 418: Meet The New DSPy: The Hot Framework to Build LLM Apps You Should Know About
Thursday, August 1, 2024
Created by Stanford researchers, the framework provides an interesting alternative to LangChain or LlamaIndex. ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
Edge 417: Building Multi Agent Systems
Tuesday, July 30, 2024
Can multiple agents collaborate in a single application? Some research and technologies that might help with that. ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
3 vs. 3: The Open vs. Closed Battle for Big AI
Sunday, July 28, 2024
Big model announcements by Meta, Mistral and xAI. ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
Edge 416: Inside Apple's 4M-21 Model that Could be the Foundation of its On-Device Multimodal Experience
Thursday, July 25, 2024
The model was trained simultaneously across 21 different modalities. ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏ ͏
You Might Also Like
🔎 How to Search Reddit Like a Pro — 9 Reasons to Always Use Windows With a VPN
Tuesday, November 12, 2024
Also: Tips for Setting Up a Mobile VR Office, and More! How-To Geek Logo November 12, 2024 Did You Know In the 2016 film Doctor Strange, the characters of both Doctor Strange and the villain Dormammu (
Web Scraping Tips, Python 3.13 Performance Boosts, Writing Interpreters & More
Tuesday, November 12, 2024
Introduction to Web Scraping With Python #655 – NOVEMBER 12, 2024 VIEW IN BROWSER The PyCoder's Weekly Logo Introduction to Web Scraping With Python In this video course, you'll learn all about
Daily Coding Problem: Problem #1606 [Easy]
Tuesday, November 12, 2024
Daily Coding Problem Good morning! Here's your coding interview problem for today. This problem was asked by PayPal. Given a binary tree, determine whether or not it is height-balanced. A height-
Charted | Breaking Down the U.S. Government's 2024 Fiscal Year 💰
Tuesday, November 12, 2024
Net interest payments cost the US government $882 billion in fiscal year 2024, the third-largest outlay in the final budget. View Online | Subscribe | Download Our App Presented by Hinrich Foundation
Spyglass Dispatch: AI's Independence Race • EU's Bad Meta Ads • AI Chip Shenanigans • Netflix Ads Religion
Tuesday, November 12, 2024
AI's Independence Race • EU's Bad Meta Ads • AI Chip Shenanigans • Netflix Ads Religion The Spyglass Dispatch is a free newsletter sent out daily on weekdays. Feel free to forward it on to
The Big T
Tuesday, November 12, 2024
Top Tech Content sent at Noon! How the world collects web data Read this email in your browser How are you, @newsletterest1? 🪐 What's happening in tech today, November 12, 2024? The HackerNoon
Deadline Extended: 2 Weeks Left to Compete for Over $7000 in the AI-chatbot Writing Contest🔥
Tuesday, November 12, 2024
Great news, newsletterest1 ! The submission deadline for the #ai-chatbot writing contest has been extended! You now have until November 21, 2024, to submit your unique AI chatbot ideas for a chance to
A very demure, very mindful issue
Tuesday, November 12, 2024
Plus a look at memory regions, Go's birthday, and we invent a brand new word. | #531 — November 12, 2024 Unsub | Web Version Together with Frontend Masters logo Go Weekly Happy Birthday, Go! Go
Visual Capitalist is revealing all of its biggest secrets... 📊
Tuesday, November 12, 2024
You can get in on our newest project if you act now. View Online | Subscribe | Download Our App We're revealing our biggest secrets... The question we get asked the most is: "How does Visual
🔓🐍 Unlock Your Python Potential with Instructor-Led Courses
Tuesday, November 12, 2024
Hey there, If you've been looking for a way to go beyond on-demand tutorials and really master Python, we've got something special for you... For the first time, Real Python is launching an