Imad Riachi

Working with some Large Language Models (LLMs) at Trionic Technologies LLP, Vatsal Patel asks a very interesting question: "Can we really exercise some influence or exploit LLMs to get focused answer from multiple choices?" On digging into it, we found a very interesting research paper that highlights the same! This paper, "Manipulating large language models to increase product visibility," shows how "strategic text sequences" could increase the visibility of the product in LLM-driven searching aids. Think about dropping a purposely designed message into a product description and seeing that one rise in the recommendation listing to the very top. Not only is it possible, but it's also now proven! Their research in optimizing these sequences employs techniques such as the Greedy Coordinate Gradient (GCG) algorithm to ensure that the very products, even if in real spaces, which were previously shrouded on digital catalogs, shine bright on the virtual shelf. This is almost revolutionary in SEO for AI-run search services in the sense of really changing the perception and priority of products by algorithms. The potential of businesses to harness this ethically is not less impactful and innovative in generating competitive edges. Let's see how strategic text sequences can help in innovations of digital marketing and e-commerce. Link: https://lnkd.in/dPE8zFtm #AI #MachineLearning #LLMs #DigitalMarketing #ecommerce

15

🆕 Benchmarks 201 with Clémentine Fourrier of the Hugging Face Open LLM Leaderboard! Why Chat Arenas are **OVERRATED**: - they bias sycophancy/assertiveness over factuality - annotators are not representative of humanity - models are argmaxxing public arena data - not reproducible - can't give fine grained test of model capabilities Why you should STOP using LLMs as Judges: - Position bias (prefer first answers) - Self bias (prefer their own answers) - Length bias (prefer long answers) - Discrete bias (can't do continuous ranges) - Reproducibility (GPT4 closed source - prefer Seungone Kim's Prometheus or Beijing Academy of Artificial Intelligence(BAAI) JudgeLM) One of our best ones 🔥

29

1 Comment

Any new significant research has to go to arXiv but everything on arXiv is really valuable? The number of AI papers on arXiv has been growing exponentially in recent years, with a doubling rate of approximately 24 months. This increase is mainly driven by the expanding scope of AI research, encompassing diverse fields such as mathematics, statistics, and physics. In 2019, over three AI preprints were submitted to arXiv per hour, highlighting the accelerated pace compared to previous decades. This surge in publications has led to challenges in keeping up with the latest advancements, prompting suggestions for AI tools to help manage and predict new research directions. I’m glad it’s summer break because most of my spare time I’ll spend catching up with pending papers 😂 ♻️ Please consider to repost if you’ve seen yourself in the same situation about arXiv! #ai #genai #research #arxiv

750

57 Comments

#AI is both over-hyped and under-hyped. Over-hyped in the sense that I don't believe artificial general intelligence (AGI) is imminent. Yann LeCun had a great interview in Wired where he outlined why (and dropped a great quote - "Machine Learning Sucks"). https://lnkd.in/exv35XGT Under-hyped in the sense that ROI is already hiding within existing business processes - someone just needs to do the work of re-thinking, re-designing and re-building with an AI first mindset. Friday, April 19th at 11:30am, join me at the River Club (https://lnkd.in/evs8XsYB) with the First Coast Tiger Bay Club for lunch, a short presentation and some q&a. To RSVP, email fctigerbay@gmail.com. The event is open to everyone, and $35 for lunch for non-members. Socrates Rivers

13

Here is Yann LeCun's (Chief AI Scientist at Meta, Prof. at NYU) take on limits of LLM's that stands in the way of these models achievening true intelligence. Whilst LLMs are great at language understanding and manipulation, he argues that they are limited in their capabilities in terms of reasoning, persistent memory, planning and understanding of the physical world via sensory inputs and interpretation (the way we humans learn), capabilities he calls essential to achieve 'true' intelligence. What are your thoughts? Share below. #LLMs #generativeAI #AGI

6

Per Yann LeCun & Demis Hassabis - the current evolution of LLMs is superficial and limited, with the models learning only when human engineers intervene to train it on that information. Agree vehemently - human reasoning is not algorithmic (read Dan Ariely etc.) and sentiments tend to be intangible not probability based (mental health, wellbeing, dating apps or divorce stats anyone?). Considering the challenges of food insecurity, equitable access to healthcare, social dissonance and societal detachment etc., why are humans trying to get #ArtificialIntelligence to 'be human-like’? "... Systems hallucinate, they don't really understand the real world. They require enormous amounts of data to reach a level of [perceived] intelligence. And they can't really reason. They can't plan anything other than things they’ve been trained on. So they're not a road towards what people call “AGI.” - Yann LeCunn, https://lnkd.in/gf2Dsfqz #Ethics, #governance, costs etc aside, I embrace the current wave of #innovation, and find the current democratization of 'AI' amazing. However, I'd advocate for the next evolution of #AGI enhancing rather than replicating human behaviour because there is nothing general about human intelligence.

17

7 Comments

AI Reading for Monday June 3! https://lnkd.in/eh-5mrsF Nvidia builds a platform for robots. - Firstpost Huang: the future is robots building robots. The robot singularity is near! - Reddit video Sam Altman borrows money to make big venture bets that pose a conflict of interest with his work at OpenAI. Bold strategy, Altman. - WSJ Jamie Dimon says JPM is up to 800 use cases for AI. - eFinancialCareers McKinsey: 12m US layoffs due to AI in next 5 years, most affected will be employees like admins, customer service, food service, and manufacturing. - Business Insider Retraining workers to use AI. - FT SEC to hold virtual meeting Thursday about AI on Wall Street. - InvestmentNews Researchers crash Baidu self-driving cars with tinfoil, cardboard and drones. - The Register Doctor says AI helped him step away from the keyboard and rekindle his love for medicine. - LA Times AI matches the quality of human phishing scams. Will need spy-vs-spy AI to keep up. - Bruce Schneier Gurman: big AI updates to Siri might not ship til early 2025; Using Apple's new OLED iPad Pro as a laptop replacement. - Bloomberg X is OK with p***y in bio if it's labelled and Musk gets his cut. Super attractive to mainstream advertisers. - TechCrunch AI safety becomes a political battlefield, and a minefield for Google and Microsoft. - Axios The Daily Caller is big mad that OpenAI doesn't want to license its high quality content. - Futurism Intel has their work cut out for them between ARM and now AMD SOCs, data center CPUs and GPUs. - Tom's Hardware AMD powers laptops with 50 TOPS of NPU and 320 TOPS of RTX4070. Won't match the battery life of the ARM laptops ofc. - Digital Trends Stable Diffusion 3 due in a couple of weeks. - The Register With LinkedInfluencer, ChatGPT will help you create posts like, "what helping an old woman cross the street taught me about B2B sales." Start influencing today! - Reddit

2

One 𝗾𝘂𝗶𝗰𝗸 method to 𝗴𝗲𝗻𝗲𝗿𝗮𝘁𝗲 𝗮𝗻 𝗶𝗻𝘀𝘁𝗿𝘂𝗰𝘁 𝗱𝗮𝘁𝗮𝘀𝗲𝘁 to fine-tune your 𝗟𝗟𝗠 ⬇ Applying LLMs to custom tasks has become more accessible since innovations like PEFT with LoRA Adapters and weights quantizations that allow faster training and low compute usage. There's a multitude of downstream tasks one could fine-tune an LLM for. Summarisation, NER (Named Entity Recognition), Instruction Following, Question Answering, etc. ⭕ When starting from scratch, dataset preparation for any of these tasks might take a very long time. 🔸 With that in mind, let's iterate over a quick method to prepare a custom dataset for the instruction-following task, similar to what the GPT-3 (no-chat) model was trained on. ↳ 𝗥𝗮𝘄 𝗗𝗮𝘁𝗮 Let's assume you want to fine-tune an LLM to inherit your writing style. For that, you'll need a set of posts/articles you've written as the knowledge base. ↳ 𝗣𝗿𝗲𝗽𝗿𝗼𝗰𝗲𝘀𝘀𝗶𝗻𝗴 You clean the samples by removing emojis, special characters, and listicles, as they might complicate the tokenization process by not fully using the token space. ↳ 𝗦𝘁𝗼𝗿𝗶𝗻𝗴 After cleaning, you usually embed and store these samples into a Vector DB (e.g QDrant) ↳ 𝗦𝗮𝗺𝗽𝗹𝗶𝗻𝗴 From the QDrant DB, you sample a subset of entries that you'll use to generate the fine-tuning dataset. ↳ 𝗘𝗻𝗴𝗶𝗻𝗲𝗲𝗿𝗶𝗻𝗴 Since our target is an instruct-based model fine-tuning, we could use a larger model (e.g. GPT3.5-Turbo) to read the sample and generate an instruction that'll describe our sample. See the example below: 🟢 RESPONSE: "Vector Databases are awesome and powerful..." 🟡 INSTRUCTION: "Write a post on why LLM enthusiasts should learn about Vector DBs" ↳ 𝗘𝘃𝗮𝗹𝘂𝗮𝘁𝗶𝗼𝗻 A good practice is to have each generated pair, logged for monitoring. One powerful tool for that is 𝘾𝙤𝙢𝙚𝙩 𝙈𝙇 𝙇𝙇𝙈 which allows you to gather and inspect prompts and chains from their dashboard. Here, HITL (Human in the Loop) analysis might be done to identify and rate the dataset preparation process. ↳ 𝗩𝗲𝗿𝘀𝗶𝗼𝗻𝗶𝗻𝗴 Once additional checks over the generated dataset are done, you can version the dataset as an artifact for future use in your fine-tuning pipeline. 🔹 This quick approach will save you a lot of time you would have spent manually composing each sample. For more in-depth details and a hands-on tutorial on how to implement it on your own, make sure to check this article: 🔗 𝘼𝙧𝙩𝙞𝙘𝙡𝙚 : 𝘛𝘩𝘦 𝘙𝘰𝘭𝘦 𝘰𝘧 𝘍𝘦𝘢𝘵𝘶𝘳𝘦 𝘚𝘵𝘰𝘳𝘦𝘴 𝘪𝘯 𝘍𝘪𝘯𝘦-𝘛𝘶𝘯𝘪𝘯𝘨 𝘓𝘓𝘔𝘴 shorturl.at/ERVy3 𝗙𝗼𝗹𝗹𝗼𝘄 𝗺𝗲 🔔 for more 𝙥𝙧𝙖𝙘𝙩𝙞𝙘𝙖𝙡 𝙜𝙪𝙞𝙙𝙚𝙨 on #machinelearning to production.

166

6 Comments

We discussed here how accurate are LLMs as automated evaluators (there are plenty of studies about this problem : how LLMs raters are aligned with human raters) Here is interesting about factual verification with LLMs. For generated texts , split the text into atomic facts (LLMs can do well, but still there is a possibility of errors), and verify facts with augmented systems having access to the knowledge bases. 2% precision will be good for many practical use cases where fact verification is needed "Our estimator decomposes generations into atomic facts and validates each based on a given knowledge source, leveraging retrieval from the given knowledge source and strong language models. Our estimator closely approximates FACTSCORE with an error rate of < 2% and can be applied to a range of new LMs at scale with no human effort. Our case study evaluates 6,500 generations from 13 LMs that could have cost $26K, with various findings: GPT-4 (OpenAI, 2023) and ChatGPT are far less factual than humans but are much better than public models, and there is a large variance between public models, with Vicuna (Chiang et al., 2023) and Alpaca (Taori et al., 2023) being some of the best" FActScore: Fine-grained Atomic Evaluation of Factual Precision in Long Form Text Generation

9

Whilst I absolutely understand the necessity for European AI legislation and also do not fully understand the EU AI Act, I think this legal framework was rushed and is going to be a hindrance in many ways and it will be incredibly difficult for European companies (big or small) too close the AI tech gap in the years to come.

5

Inexpensive token generation and agentic workflows introduce fresh prospects for training Large Language Models (LLMs) with synthetic data, potentially elevating output quality. Direct pretraining of LLMs on their generated responses proves ineffective, but employing agentic workflows could yield superior results, akin to humans learning from their own thinking processes. LLM training typically comprises pretraining on vast amounts of unlabeled text data, followed by fine-tuning and alignment to human values. However, sourcing adequate training data remains a hurdle, prompting exploration of innovative methods like training smaller models on the output of larger ones. Despite the substantial cost of token generation, agentic workflows present promising avenues for creating high-quality synthetic data, ultimately enhancing LLM performance.

11

🩺 Quantum-classical Bayesian neural networks are not only long enough to be a title all their own, but taking on a challenging medical dataset. Talk about ambitious. Today's dispatches from quantum: ⚆ Zaiku Group Ltd, National Quantum Computing Centre (NQCC), and Oxford Quantum Circuits (OQC) are using hybrid classical-quantum federated learning to securely handle massive biomedical datasets, showing promise in genetic data analysis ⚆ MITRE advises the next U.S. administration to prepare for quantum computing's potential to crack encryption, emphasizing advancements in infrastructure protection and a zero trust framework ⚆ Berkeley Lab uses femtosecond lasers to create precise telecom-band optical qubits in silicon ⚆ Riverlane joins Tech Nation’s Future Fifty to develop a quantum error correction stack for error-free operations ⚆ Institute for Quantum Computing team completes a quantum source for the QEYSSat, Canada’s first quantum satellite ⚆ IonQ and SC Quantum kick off a Faculty & Researcher Learning Series to promote quantum tech adoption in academia and business ⚆ Quantum phase estimation and quantum algorithms for the discovery of homogeneous catalysts, improve design and efficiency in chemical manufacturing ⚆ A proposed quantum-classical Bayesian neural network improves reliability in medical image classification by combining classical CNNs with quantum circuits for better uncertainty estimation ...and more. 📰 Stay quantum-savvy. The Daily Qubit offers daily insights into the rapidly evolving world of quantum computing. Join leaders in academia, founders, and industry experts in staying informed. Sign up at www.thedailyqubit.com and never miss a qubit. #quantumcomputing #quantumphysics #innovation

3

Fascinating Low Valuation on IPA Therapeutics(NASDAQ:IPA $1.29) On March 29, 2022 IPA paid $20 million pounds or approximately $26.4 million USD for BioStrand AI. The pound was trading(1 pound = 1.32 USD) at this time. Since this acquisition BioStrand AI has announced several significant events/successes including, granting of the HYFT Technology Patent by the European Patent Office(EPO), BioStrand solves the Information Integration Dilemma(IID), BioStrand creates a revolutionary approach to biological sequence retrieval, BioStrand introduces an advanced Foundational AI Model utilizing LLM stacking & HYFT Technology, InterSystems & BioStrand AI Collaborate to Unveil Innovative Integration of Vector Search, BioStrand's AI has been learning from IPA's leading, proprietary Anti-Body Discovery Database(possibly the best in the World)for the last 24 months, and much more. IPA has been working on antibody research for 40+ years. https://lnkd.in/gJHWdgck https://blog.biostrand.ai Yet, the entire current Market Capitalization of IPA(CRO + BioStrand + Talem) is approximately $35 million USD today. IPA's Contract Research Organization(CRO) has had 3 record revenue quarters in a row, is growing at approx. 20%, and Fiscal 2024 Revenue should be around $20 million USD. They deal with 19 of the top 20 pharmas, and have over 600 partners. One could argue that the CRO is worth between $80-$120 million. FairJourney Biologics(an excellent CRO) was just put up for sale for $1 Billion+. Although FJBio has $45 million in EBITA, IPA has better numbers regarding projects, libraries generated, partners, species used, and 1/2 the employees. In other words, better data to train the AI models. Wet lab combined with AI will be critical moving forward. I also believe BioStrand AI is undiscovered by most stock market participants. For Information Purposes Only. Please complete your own due diligence. CFP has a position in IPA. Is a short squeeze possible? Sum of the parts(1 + 1 = 11) IPA (ImmunoPrecise Antibodies)BioStrand (a subsidiary of IPA) InterSystems #ai #HYFTTechnology #drugdiscovery #LENSai #antibodies #LLMS #NLP #semantics #pharma #healthcare #biotech #multimodality

2

Yann LeCun advises students entering the AI field to avoid working on large language models (LLMs). Speaking at VivaTech in Paris, he emphasized that developing next-generation AI systems that overcome LLM limitations would be more impactful. LeCun highlighted the need for innovation in AI architectures beyond LLMs and warned against the dominance of a few AI assistants in controlling digital interactions, stressing the importance of diversity and open-source development in AI. #YannLeCun #AIAdvice #VivaTech #NextGenAI #AIInnovation #BeyondLLMs #AIResearch #OpenSourceAI #AIDiversity #FutureOfAI #TechLeadership

1

Aporia today released its 2024 Guardrails Benchmark report, highlighting the performance of Aporia in achieving industry-leading recall rates and markedly lower latency when compared with NVIDIA NeMo, GPT-4o, and GPT 3.5. Results are indicative of Aporia’s commitment to advancing #AI deployment standards, providing organizations and development teams with a trusted solution for releasing #AIapplications that are both responsive and secure. “Aporia is focused on empowering engineers and organizations to deploy secure and reliable AI applications without compromising on performance,” said Liran Hason, CEO and Co-Founder of Aporia. “We take pride in these benchmark results, which highlight the efficiency and accuracy of our solution and showcase the company’s commitment to enhancing AI reliability. In the ever-evolving AI industry, we acknowledge the ongoing need for improvements and are dedicated to continually raising the bar for AI safety and performance.” Read more here: https://ow.ly/xB6O50SELvm

5

2 Comments

If you want to understand why LLMs are not the solution for AGI (Artificial General Intelligence), I highly recommend listening to Lex Fridman podcast with Yann LeCun : Meta AI, Open Source, Limits of LLMs, AGI & the Future of AI T͟L͟D͟R͟: 1. LLMs do not understand the physical world 2. LLMs have not the ability to remember 3. LLMs have no persistent memory 4. LLMs are not able to reason and to plan

27

5 Comments

Launching a Pioneering Initiative at coeo AI Lab for Quantum-Inspired Transformer Models 🚀 We're thrilled to announce the commencement of our cutting-edge initiative aimed at exploring and developing novel approaches in the realm of Transformer-based Large Language Models (LLMs), drawing inspiration from quantum mechanics. As we delve into the initial conception and brainstorming phase, our mission is crystal clear: harness quantum-mechanical analogies to forge more efficient algorithms for word embeddings and attention mechanisms, tailored to our unique use cases and datasets. 🔬 Our primary objective is to intertwine methods from many-particle quantum mechanics with classical machine learning paradigms, enhancing their power and efficiency. Beyond that, we envision creating a synergy with Quantum Computing by translating concepts from classical Transformer architectures into a versatile quantum-mechanical framework. In our dedicated AI Lab, theoretical physicists, data scientists, and AI engineers collaborate closely, transforming groundbreaking scientific insights into practical prototypes. But our ambition doesn't stop there. 🌐 We aim to foster cross-industry collaborations with partners from academia and the business sector. By doing so, we're not just broadening our perspective but also unlocking a vast array of potential applications. Stay tuned for updates as we progress, and we look forward to engaging with the scientific and industrial community for a vibrant exchange of ideas. #QuantumComputing #MachineLearning #AIResearch #Innovation #Collaboration #DataScience #AI #genAI #generativeAI #artificialintelligence #coeo #CoeoGroup

25