Aman Gupta

A recent breakthrough titled "Matrix Multiplication-Free LLMs" demonstrates a huge advancement in the area of Large Language Models (LLMs) by reducing computational costs. The authors have eliminated MatMul operations from LLMs, claiming to 10 times reduction in memory usage and a 25.6% increase in training speed, all while maintaining strong performance at billion-parameter scales. Paper link: https://lnkd.in/ggph8qXc #AI #machinelearning #deeplearning #LLMs

16

AI at Meta has released the most powerful open-source model yet as of today: 𝑳𝑳𝑨𝑴𝑨3 - 8B and 70B models at 8k context length! They highlight improvements in each of the following aspects as the key differentiator: a) Model Architecture b) Pretraining data c) scaling up pre-training d) Instruction fine-tuning 💥 𝐏𝐫𝐞𝐭𝐫𝐚𝐢𝐧𝐢𝐧𝐠 & 𝐒𝐜𝐚𝐥𝐢𝐧𝐠 𝐮𝐩: ✈️ Vocab of 128k tokens & GQA across both 8B and 70B, 8k native context length ✈️ Pretrained over 15T tokens, 4 times more code than LLAMA2 ✈️ Multi-lingual : over 5% of pretraining data consists of high-quality non English data that covers over 30 languages: don't expect same performance as English ✈️ LLAMA2 generates training data (like in Self-Instruct) to generate text classifiers that filter data powering LLAMA3 ✈️ Both 8B and 70B improve even after the model is trained on 2 orders of magnitude more data i.e log-linear improvement after 15T tokens training ✈️ Training runs on 2 custom-built 24k GPU clusters with an overall training time efficiency of 95% ✈️ 𝘖𝘷𝘦𝘳𝘢𝘭𝘭, 3 𝘵𝘪𝘮𝘦𝘴 𝘮𝘰𝘳𝘦 𝘦𝘧𝘧𝘪𝘤𝘪𝘦𝘯𝘵 𝘵𝘩𝘢𝘯 𝘓𝘓��𝘔𝘈2 💥 𝑰𝒏𝒔𝒕𝒓𝒖𝒄𝒕𝒊𝒐𝒏 𝑭𝒊𝒏𝒆𝒕𝒖𝒏𝒊𝒏𝒈: ✈️ Combination of SFT + PPO + DPO + Rejection Sampling ✈️ 𝑸𝒖𝒂𝒍𝒊𝒕𝒚 𝒐𝒇 𝒑𝒓𝒐𝒎𝒑𝒕𝒔 𝒖𝒔𝒆𝒅 𝒊𝒏 𝑺𝑭𝑻, 𝒑𝒓𝒆𝒇𝒆𝒓𝒆𝒏𝒄𝒆 𝒓𝒂𝒏𝒌𝒊𝒏𝒈𝒔 𝒖𝒔𝒆𝒅 𝒊𝒏 𝑫𝑷𝑶, 𝑷𝑷𝑶 𝒉𝒂𝒔 𝒂𝒏 𝒐𝒖𝒕𝒔𝒊𝒛𝒆𝒅 𝒊𝒎𝒑𝒂𝒄𝒕 𝒐𝒏 𝒎𝒐𝒅𝒆𝒍 𝒑𝒆𝒓𝒇𝒐𝒓𝒎𝒂𝒏𝒄𝒆 ✈️ Careful data curation and multiple rounds of QA on annotations from human annotators ✈️Training on preference ranking improved the model performance greatly on coding and reasoning tasks 💥 𝑳𝑳𝑨𝑴𝑨3 𝑮𝒂𝒖𝒓𝒅𝒓𝒂𝒊𝒍𝒔: ✈️ Updated and new safety tools : LLAMA Gaurd2 & cybersecEval 2 ✈️ CodeShield - inference time gaurdrail for filtering insecure code ✈️ Instruction Fine-tuned model redteamed for safety : By generating adverserial prompts that try to elicit problematic responses. ✈️ LLAMAGaurd - foundational for prompt and response safety, can be easily finetuned to create a new taxonomy 💥 𝑰𝒏𝒇𝒆𝒓𝒆𝒏𝒄𝒆 & 𝑭𝒖𝒕𝒖𝒓𝒆 ✈️ Despite LLAMA3 having 1B more params than LLAMA2 7B, the improved tokenizer efficiency and GQA contribute to maintaining the inference efficiency on par with Llama 2 7B. ✈️ 400B+ parameters, Multilingual, Multimodal, and longer context window LLAMA3 would be available on AWS soon. Try it out! #generativeAI #llm #llama3 #aws #bedrock #sota

33

1 Comment

1. IMO, agent patterns are one of the most interesting ways to extend LLMs. memary is based on leveraging a graph database (neo4j) for long-term memory (to get around the limited memory inherent in even the largest context windows in LLMs) and augmenting it with other tools like web search and Computer Vision. https://lnkd.in/gfFcBqJx Looks interesting. 2. More about how consolidation effects the entire healthcare industry: https://lnkd.in/gXJKqNJB The Change hack/ransomware attack not only affected claims, service, refills, etc. but now CMS & NCQA are delaying the entire HEDIS reporting process (used for Medicare/Medicaid quality ratings) due to the impact. 3. NARA bans ChatGPT use agency-wide. https://lnkd.in/gW8T94JU Federal Government agencies see this as a risk, including their information leaking and being used to train models (which may then output the wrong information or show PII/PHI other proprietary information). 4. Oh great, now I can't even be angry anymore :) https://lnkd.in/gcrN9mjx

1

1 Comment

Hallucinations in LLMs are inevitable, no seriously! Josh Fritzsche and I have worked hard at Voze to find a way to tune, prompt and even plead 🙏 with LLMs to prevent them from the occasional 🍄 hallucination. It seemed like the better we have done the harder the edge cases have been to solve and eventually we realized that they tend to frequently happen when LLMs predict or make anticipations on "nothing". I remember at one point nearly going insane with frustration because when LLMs had a hallucination, they were EPIC! 😱 Then we pushed through it and benefited for the effort! How did we solve this common problem? We didn't (Not completely), we got better at identifying the issues as they happened. Once we identified in the data we were able to start looking for patterns. The same catalysts that cause it to happen once if left unchanged caused it over and over again. Our solution is astoundingly simple. Once you know that it's likely to happen..... let it! Accept it will 🍄, but also be preemptive in not focusing on it and use alternative data sources, generators or providers. Change the conditions that cause it but continue to monitor it as it happens. As Dan Caffee is fond of saying to me on a regular bases. "Dusty we need the system to be adaptive and provide mechanisms to learn from it's past." Well he is right! All we needed was data and the people that contribute to it's quality. Huge props 👏 to strong support from Kathy and Janelles audit teams. We were armed with a knowledge and a desire to learn from past mistakes. Armed with confidence that we could correct these problems we could focus on the data to prevent their impact going forward. When doing millions of LLM (Generative AI) calls regularly Josh and I have learned that hallucinations will happen. We also know exactly how we will handle and prevent it going forward! Failure is not doing anything and accepting defeat.

4

2 Comments

*Mixture of Agents (MoA)* is an innovative way to boost Large Language Models’ performance by leveraging the strengths of multiple agents as a team. Layered Architecture: MoA arranges many LLMs in layered architecture, each layer composed of numerous individual agents(LLMs). Collaboration: These agents generate responses that are based on outputs from the previous layer’s agents. They repeatedly refine and enhance the last output. Diverse Insights: By integrating different capabilities and insights from various models, MoA achieves a more powerful combined model that is secure against diverse risks. Performance Boost: MoA outperforms standalone LLMs by a wide margin. For instance, it set a new benchmark with 65.1% on AlpacaEval 2.0 as against GPT-4 Omni at 57.5%, using only open-source models. In short, if we combine language modelers’ strengths then it results in increased efficiency and versatility. Here is a nice implementation of MOA by Together AI Github source: https://lnkd.in/dMCVRS3D #AI #LLM #LargeLanguageModels #MachineLearning

59

3 Comments

🔥 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈 𝐚𝐧𝐝 𝐋𝐋𝐌 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 - 𝐓𝐨𝐩 𝐪𝐮𝐞𝐬𝐭𝐢𝐨𝐧𝐬 🔥 🤖 I've conducted a mock interview with my student, Bheesetti Anand, focused on Generative AI and LLM (Large Language Models) interview questions. This session is designed to benefit other students preparing for similar interviews, helping them secure their dream roles in the field of Generative AI. 📔 Here’s why watching the video can help you make a 𝐬𝐮𝐜𝐜𝐞𝐬𝐬𝐟𝐮𝐥 𝐜𝐚𝐫𝐞𝐞𝐫 𝐬𝐰𝐢𝐭𝐜𝐡 𝐭𝐨 𝐆𝐞𝐧𝐞𝐫𝐚𝐭𝐢𝐯𝐞 𝐀𝐈: ✅ 𝐂𝐨𝐦𝐩𝐫𝐞𝐡𝐞𝐧𝐬𝐢𝐯𝐞 𝐂𝐨𝐯𝐞𝐫𝐚𝐠𝐞: Detailed exploration of top interview questions related to Generative AI and Large Language Models. ✅ 𝐏𝐫𝐚𝐜𝐭𝐢𝐜𝐚𝐥 𝐈𝐧𝐬𝐢𝐠𝐡𝐭𝐬: Real-world examples and practical applications to understand complex concepts better. ✅ 𝐄𝐱𝐩𝐞𝐫𝐭 𝐆𝐮𝐢𝐝𝐚𝐧𝐜𝐞: Tips and strategies to approach and answer interview questions effectively. ✅ 𝐂𝐚𝐫𝐞𝐞𝐫 𝐏𝐫𝐞𝐩𝐚𝐫𝐚𝐭𝐢𝐨𝐧: Essential knowledge and skills to prepare for interviews and stand out to potential employers. 📽 𝐢𝐧𝐭𝐞𝐫𝐯𝐢𝐞𝐰 𝐥𝐢𝐧𝐤: https://lnkd.in/g_nkP7fs #generativeai #llm #interview #jobpreparation #mockinterview

14

2 Comments

These findings motivate our design of a novel framework to investigate hidden representations in LLMs with LLMs, which we call Patchscopes. The key idea behind this framework is to use LLMs to provide natural language explanations of their own internal hidden representations. Patchscopes unifies and extends a broad range of existing interpretability techniques, and it enables answering questions that were difficult or impossible before. For example, it offers insights into how an LLM's hidden representations capture nuances of meaning in the model's input, making it easier to fix certain types of reasoning errors. While we initially focus the application of Patchscopes to the natural language domain and the autoregressive Transformer model family, its potential applications are broader. For example, we are excited about its applications to detection and correction of model hallucinations, the exploration of multimodal (image and text) representations, and the investigation of how models build their predictions in more complex scenarios. Google DeepMind

8

Interesting announcement from Hugging Face about ORPO. My key take away is that a lot of energy is spent on making fine tuning of LLMs even more simple and accessible. - ORPO is a new, efficient fine-tuning technique for LLMs that combines supervised learning and preference alignment. - Llama 3 enhancements improve performance with increased context length and a new tokenizer. - It is easy to access ORPO fine-tuning code for developers and researchers. Read the great blob. Thanks Hugging Face https://lnkd.in/eS74uG79

16

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

Wondering what top research got published this last week? 📑🤖🤔 So essentially, 1. "AI Model StyleFeatureEditor gives highly detailed image editing results!" 2. "We have a Universal Quantum Chemistry dataset to find new drugs now!" 3. "Stylebreeder clusters diverse art styles from 6.8M images on Artbreeder, enhancing personalized AI-driven artistic expression." 4. "LongRAG addresses the limitations of traditional RAG by using long retrieval units" 5. "Language model tuned to avoid toxic language in English will also avoid it in other languages!" 6. "YOUDREAM Auto-Generates Perfect 3D Creatures from Text!" From cutting-edge research on Artificial Intelligence to advances in Machine Learning, Computer Vision, and Neuroscience, we distill the latest technical papers into bite-sized insights you can apply today! For more subscribe at https://lnkd.in/gJ_wJxAC

21

1 Comment

https://lnkd.in/dJW9sZpx SelfExtend introduces a novel approach to expanding the context window of LLMs. It leverages a two-tiered attention mechanism—grouped attention for long-range dependencies and neighbor attention for local interactions—both derived from the model's original self-attention. This method allows seamless context window extension without requiring fine-tuning, and extensive experiments have demonstrated its effectiveness across various benchmarks.

12

1 Comment

A couple of recent college graduates from India against AI finetuning giants like Anyscale and Run:ai? I met Arko C at a NASSCOM event last October. He was building Xylem AI (YC S24) , a platform that allows you to train and deploy LLMs in production. So naturally most Indian VCs rejected him. Because how could a team of three recent college grads compete with the likes SV startups like Anyscale that have raised hundreds of millions of dollars, have teams of experienced AI researchers and a stash of H100 GPUs? Arko hustled and closed two Fortune 500 companies for paid contracts. I personally know unicorns that haven't managed to do this. He then got into YC in the last round in May. There's a lot to still be proved out, but Arko exemplifies how valuable grit is in building startups. I’m doing a webinar with him tomorrow on challenges of using LLMs in production. Link to register in the comments. Disclaimer: I'm a small investor in Xylem.

153

5 Comments

We just released a blog on whether the use of multimodal feature transfer can detect deception (phenomena like sarcasm, irony, and condescension) in conversation. The short answer... Yes. This particular write-up profiles a subset of our work at Symbl.ai on multimodal large language models, and is research led by one of our summer interns, Benjamin R. (from Prof. Larry Heck's AVA Lab at Georgia Institute of Technology). We have a paper currently under submission on this work, and we can't wait to share the full details with all of you soon. Blog Link: https://lnkd.in/g6m_wFDi #multimodal #featuretransfer #LLMs #deception #conversation #sarcasm #irony #language #AI #ML #NLP #LargeLanguageModels #SummerInternship #ResearchPapers #ACL #Llama2 #SymblDotAI

16

Had a blast giving an online talk at an Indeed conference on "Best practices of ChatGPT for NLP applications: Data Labeling, Data Generation and Benchmarking" two weeks back. Thanks Madhu Kurup for the opportunity, I had a lot of fun! 😊 The field of LLMs is so hyper-rapidly evolving, that it is important to pro-actively keep abreast with the field, so we can connect the dots and customize it for industry applications and business use-cases. The talk attempts to deliver one such perspective from my experience and understanding of the field. YouTube link: https://lnkd.in/geY8e8ht #llms #largelanguagemodels #genai #generativeai

25

While being futuristic, Quantization of a Neural Network model is the key to reduce dependency on high end hardware. This one is a great explanation of mode quantization techniques that can save millions of 💵 trying to implement LLM #LLM #ConversationalAI #GenAI #Chatbots #contactcenter

7

New post! Probably the most desired enhancement for database privilege management is finally here in 23ai. #database #sql https://lnkd.in/gXkXGWWH

31

⚡ Accelerating the inference of LLMs ⚡ One of the benefits of making technology open-source is that the community starts to research ways to use it at lower costs (because not all of us have good infrastructure at home). This is a key point on LLMs as they are massive neural networks that need strong GPUs to work. Since Meta released Llama (one of the first LLMs to have its weights available) many techniques have been raised to let people use it on their computers. One famous example of this is Lora, a technique to fine-tune LLMs with a small computing cost. A new work in this sense is DeJavu (https://lnkd.in/dvizeV_Y) a technique to speed up the model's perdiction. The idea is straightforward, train another neural network to predict which neurons of each layer will be useful for a given input. Then, both networks run in parallel letting the LLM use only important neurons of that particular case. The technique reduces computing time to half or less and maintains the output's quality. I think this work brings two important insights: 1) Models are huge but not all of their neurons are needed at the same time 2) Open-sourcing technology is the best way to progress ❤️ As the famous proverb says (and ensemble techniques have demonstrated): Many heads are better than one #IA #LLM #OPENSOURCE

8

https://lnkd.in/esZgE24d Seven years ago I gave this talk at IIT Kanpur. Back then it seemed inevitable that we would have self-driving cars by now, and some predictions I made were wrong. Overall, though, I think this talk has aged reasonably well. Little did I know back then that the reference to the 2016 MaxSAT I gave would spark the creation of a new company, InsideOpt. The Seeker Solver that we sell is based on the technology outlined, and it is changing the optimization landscape. On the social implications, we since edited a special issue of the AI Magazine on the social disruption that AI brings. More work is needed. We need to have a wide societal discourse on how we are going to share the increasing wealth that AI brings to the world. #insideopt.com

8

Going through the list of papers for #ICML2024 and #ICLR2024 and found some interesting ones that I'll share over the next few days. But here's the big one that got me hooked: Any-Precision LLM: Low-Cost Deployment of Multiple, Different-Sized LLMs Paper: https://lnkd.in/dWkn_d_6 Authors: Yeonhong Park and others. Essentially, the paper proposes: ➡️ Rapid generation of multiple models at different quantization levels → Think about having 2, 3, 4, 5, 6, 7 or 8-bit models always in memory, but with a memory requirement of only an 8-bit model (well, almost, they do need some metadata but its an order of magnitude lower memory requirement). ➡️ Specialized LLM serving engine with the support to serve different quantization levels of the same model. ❓ Why is this important ❓ Imagine the scenario where you have a highly finetuned LLM (or an SLM) that you deploy for your users. As a failsafe, you also need to deploy a secondary LLM that comes into play if the first one goes down. However, as soon as the secondary LLM comes online, the users will notice - it's not a finetuned version of your first model! Or even if you deploy a finetuned version, you end up deploying TWO models, in memory, in parallel. This paper solves this problem - a possible failsafe method for an in-production LLM that failovers to smaller/higher quantization levels based on bandwidth and user requirements. Incredible! For a truly redundant system though this still doesn't work because if the LLM serving engine goes down, all models go down as well. So don't throw away your failsafe LLM just yet, but this might be a handy alternative. Book a free 15 minute call with me: https://lnkd.in/dGnHs_Mn #icml2024 #iclr2024 #llm #nlp #machinelearning #consulting #generativeai #openai #microsoft #iclr #icml #research #paper

12