Jonathan Roberts

Anthropic has launched Claude 3.5 Sonnet, the inaugural release in their upcoming Claude 3.5 model family. This new AI model sets unprecedented industry benchmarks for intelligence, outperforming competitor models and even Anthropic's own Claude 3 Opus across a wide range of evaluations, including graduate-level reasoning, undergraduate-level knowledge, and coding proficiency. Claude 3.5 Sonnet demonstrates significant improvements in understanding nuance, humor, and complex instructions, while excelling at producing high-quality content with a natural, relatable tone. ➜ Check the news at: https://lnkd.in/dbKMKPm7 ➜ Vendorvise's official Instagram: https://lnkd.in/ebK2KQ2C ➜ Vendorvise's Linktree: https://lnkd.in/dQdbMdth 🤖 If you enjoy reading my posts about AI and tech news, hit that like button and share with your friends to support my content! #ai #artificialintelligence #ainews #aipowered #news #blog #claude3 #sonnet #anthropic #generativeai #llms #tokens #aitools #aitech #tech #technology #hardware #content #socialmedia #future #new #launch #business #techtrends #innovation #success #markets #people #industry

2

Very insightful work by Anthropic’s interpretability team. And an amazing paper, with outstanding writing and figures. The idea is very simple: interpret LLMs by leveraging sparse autoencoders as surrogate models of the MLP of transformer blocks, which allow one to disambiguate the superposition of features captured by a single neuron. A simple idea, but a very careful and complex execution, as it is often the case in our line of work. The paper goes into many details and provide a large array of insights, although the gist of the implementation remains obfuscated due to the closed source nature of Claude. Too bad, because this is the kind of work that we need to better understand and eventually trust LLMs. This is demonstrated by the authors in the section ‘Influence on Behavior’, where they show that clamping some features to either high or low value during inference is “remarkably effective at modifying model outputs in specific, interpretable ways”. Hopefully this kind of work is going to be replicated and generalized to open-weights models, such that we have new ways to steer their behavior. https://lnkd.in/eVym7f_f #interpretability #xai #explainableai #steerableai #anthropic #claude #anthropic

3

Super glad to have contributed a small part to the updated Gemini 1.5 report in measuring in context planning capabilities https://t.co/Pltp92WcNo with Huaixiu Steven Zheng Swaroop Mishra Hugh Zhang Xinyun Chen and Azade Nova

122

3 Comments

Deciphering AI's Inner Workings: Scaling Sparse Autoencoders to Claude 3 Sonnet - A good read! https://lnkd.in/eYuwg7_U

9

🔍 The Future of AI with OpenRLHF 🔍 As AI technology advances, aligning large language models (LLMs) with human values and intentions becomes increasingly critical. Reinforcement Learning from Human Feedback (RLHF) is a powerful technique addressing this challenge. However, traditional RLHF frameworks struggle with the complexity and resource demands of training models exceeding 70 billion parameters. Launching this week is **OpenRLHF** is an open-source framework designed to overcome these limitations by leveraging cutting-edge technologies such as Ray, vLLM, and DeepSpeed. Here's why OpenRLHF is a game-changer for AI development and why I am so excited! 1. Scalability - OpenRLHF efficiently scales RLHF training for LLMs beyond 70 billion parameters by distributing models across multiple GPUs, optimizing memory use, and minimizing computational overhead. 2. Performance Optimization - By integrating Ray for model scheduling, vLLM for accelerated sample generation, and DeepSpeed for enhanced training, OpenRLHF ensures superior performance and reduced training time. 3. Versatility and Usability - Fully compatible with the Hugging Face library, OpenRLHF supports various alignment techniques like Direct Preference Optimization (DPO), Kahneman-Tversky Optimization (KTO), and rejection sampling. It offers user-friendly, one-click training scripts for diverse models and algorithms. 4. Resource Efficiency - OpenRLHF’s scheduling and memory management reduce GPU memory fragmentation and communication overhead, enabling larger batch sizes and more efficient training processes. OpenRLHF sets a new standard for AI development and ResponsibleAI. Explore the full potential of OpenRLHF and its strategic benefits in the detailed report. Read more https://lnkd.in/eCF7Q9Pb #Innovation #AI #Leadership #RLHF #OpenSource #TechAdvancement #AILeadership #FutureOfAI

5

Coming across a misleading photo, an incendiary Tweet or a comment chain full of bots and trolls is very common whenever we're on the Internet. Especially when "doomscrolling", it's quite easy for us to overlook misinformation on social media, or be influenced by it. To combat this, USC Information Sciences Institute researchers have developed machine learning models to identify coordinated inauthentic behaviors by accounts on X involved in influence campaigns around the world. Check out my article about this: https://bit.ly/3QNOOa6 #research #socialmedia #machinelearning #AI #misinformation #influencecampaigns

14

1 Comment

I decided to see if Claude 3.5 Sonnet is as good as Anthropic claims. I came up with 5 tests to pit it against OpenAI's GPT-4o. 1. Handwriting: My handwriting is terrible so I wrote my prompt in a notebook. https://lnkd.in/d-q4Rprj 2. Making a complete game: Using Python create a tower defense game from scratch. https://lnkd.in/d7ZaiyP8 3. Vector art: Make me a cool spaceship logo. https://lnkd.in/dbxQN3Jn 4. A funny story: Write a tale about a humorous cat. https://lnkd.in/dAyrmT8x 5. The big debate: Should AI have rights? https://lnkd.in/dK4EG7GE

8

2 Comments

If you were to design a statistical problem in a lab, I doubt you could come up with anything more challenging than media mix modeling. So, how do we tackle it here at Recast? Statistical research is at the core of what we do, and it’s reflected across our organization: over 30% of our team has a PhD in either math or statistics, and I wouldn't be surprised if we had the most complex Bayesian model in the world. We take a lot of pride in our model, being research-focused, compute-heavy, and having a great team of scientists. But one of the challenges in media mix modeling is that you’re communicating to marketers, and they have to use the model to make day-to-day decisions about where they're going to allocate spend. And not just that – they also need to be able to communicate their rationale to their own stakeholders and executive teams. So, while we’re scientists, they’re often not. And so the results of our model need to be communicated in an accessible way that is easy to understand, use, and relay within their organization. Not only do we have to be the best in the world at creating a statistical computing engine to estimate models of this complexity, but we also have to break it down into digestible insights and recommendations. We work very closely with our customers every day to make sure that the information that we're learning about their marketing mix is being communicated effectively and that they can understand the work that we're doing. That’s why we say Recast is built *by* scientists *for* marketers.

31

2 Comments

#RAG DOES NOT REDUCE HALLUCINATIONS IN #LLMS: Think I've called this out a few times, but the validation is nice - https://lnkd.in/eAi6zxW2. The article makes two well established points: 1) "Hallucination is a problem of reasoning and not relevance! Any amount of relevant text fed through RAG to a machine will retain the original perplexity and entropy in the system to hallucinate independent of the text"; and 2) "The initial challenge arises from the retrieval phase’s reliance on semantic similarity, which is quantified by inner product spaces or cosine similarities. This metric, while capturing semantic closeness, does not differentiate between factually accurate and inaccurate information." This should be clear to everyone, but the industry still shills flimsy RAG scaffolding anyway (shovels to gold miners).

104

34 Comments

Loved the talk. I know category theory but never saw a software implementation side-by-side comparison as he does here using Python and Topos Institute's AlgebraicJulia. Highly recommend it to those that want to think about software engineering in a different way. His use of the triangle graph definitely reminds me of my current implementation of algebraic evaluation in the ntqr Python package. I know the implementation is brittle (coded for exactly three noisy classifiers) and getting the pieces to work has had its issues. I'm definitely going to think about how to do a general n-ensemble implementation using AlgebraicJulia.

11

1 Comment

Considerations for AI Opt-Out https://buff.ly/3JxDFpG Creating a Large Language Model (LLM) requires a lot of data – as implied by the name, LLMs need voluminous input data to be able to function well. Much of that content comes from the Internet, and early models have been seeded by crawling the whole Web. This now widespread practice of ingestion without consent is contentious, to put it mildly. Content creators feel that they should be compensated for providing this input data, or at least have a choice about whether it is used; AI advocates caution that without easy access to input data, their ability to innovate will be severely limited. Source: Considerations for AI Opt-Out ()

2

I’m thrilled to announce the launch of my new Substack, Silicon Sand Stack, with my very first article: "Today's Large Language Models Will Never Be Responsible, Safe, or Green: The Need for Investment in Revolutionary New AI Paradigms" In this debut piece, I dig into the critical flaws of current large language models (LLMs) and why they fall short of being responsible, safe, or environmentally friendly. I examine the inherent issues like hallucinations, lack of real-world grounding, unreliable reasoning, and potential biases that plague these models. 🔍 Key insights include: - The technical and ethical limitations of today's LLMs. - The necessity for radically new AI paradigms. - The role of venture capital in fostering innovative AI approaches that prioritize safety, responsibility, and sustainability. Join us as we embark on a journey to rethink AI and advocate for groundbreaking advancements that align with human values and societal needs. Read the full article and subscribe here: https://lnkd.in/eDcTniTg #ArtificialIntelligence #AI #GenerativeAI #TechInnovation #ResponsibleAI #GreenAI #FutureOfAI #SiliconSandStudio

46

1 Comment

A little preview of a tool I built with Pyviz to analyze a first-stage knowledge graph. I have other plans for this, but #AI #knowledge - This helps me to visualize edge + and node relations among other items, from a core sampling model I created to generate the nodes and edges. #LLM #RAG #datascience

2

1 Comment

In computational linguistics, researchers are focused on enhancing language models' ability to extract precise information from extensive textual data. Recent studies introduce "retrieval heads," specialized attention mechanisms that selectively focus on critical segments of large texts, improving targeted data retrieval. These heads, integrated into transformer-based models like LLaMA and Yi, significantly enhance accuracy and efficiency, particularly in handling long-context scenarios. The methodology involves thorough experiments across various models, including the Needle-in-a-Haystack test, which embeds specific information to measure precision. Results show models with active retrieval heads outperform those without, with accuracy dropping notably when these heads are masked. The empirical evidence underscores the effectiveness of retrieval heads in improving information retrieval precision and reliability. This research deepens our understanding of attention mechanisms in large-scale text processing, paving the way for more efficient and accurate language models. https://lnkd.in/dNNPvMXw https://lnkd.in/ddbHX6EA Innovation Hacks AI #aiinnovation #generativeai

1