Viral Gupta

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

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 𝘓𝘓𝘈𝘔𝘈2 💥 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

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

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

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

New iterative reasoning technique could improve RAG app accuracy by 50% (vs. Naive RAG)! Another solid example of the rapid innovation cycle on RAG research. BTW, you can quickly find out answers to question like "Why did combining RAG with Fine-Tuning and Iterative Reasoning, such as OODA, improve accuracy by 50 percent?" in Epsilla (YC S23)'s "RAG Research Paper" smart search app at: https://lnkd.in/gVgzpbue , with number of research papers included exceeding 360. #genai #rag #smartsearch #innoation

18

Excellent essay about LLMs and their limitations. This is covered in a fascinating, in-depth in a post from Rohit Krishnan (direct link to post in 1st comment). Rohit write this up on on strangeloopcanon .com What's interesting is despite how capable LLMs seem from executing tasks in coding, "art" creation, and simply answering questions, LLMs often struggle with seemingly simple games like Wordle and Sudoku. Rohit describes this pointing to inherent design limitations, like "goal drift," where models lose focus over iterative tasks. I think it's great (as a user) to start understanding these weaknesses to help better grasp the boundaries of AI capabilities and how to best use this technology in it's current state. #ai #agi #artificialintelligence #openai ---- Also, I'm trying something a little weird here and putting the link to the article in the comments to see if the post gets more impressions. It seems that any post with an external link is heavily throttled while even the silliest and most useless posts without external links pop off. So, check the comments for the link to the original article.

3

1 Comment

What are your thoughts on heavily using LLMs to create technical documentation? General idea is that there is a lot of (open) technical documentation, and given LLM's good coding knowledge - it probably has a good mapping system between code and (good) docs.

10

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

‼️ Go team! 🙌 tldr: AI at Meta has shown that instead of choosing single token prediction teams can opt for multi-token prediction (which would essentially be multi-threaded processing...) The ROI is 3 x [old_ROI] wrt (inference) performance with $0.000 additional costs and/or performance degradation during the (pre)training process! 🤯 #llm #ai_innovation #ai Meta does acknowledge teams have used this approach for #finetuning. Therefore, the approach is not new (per se); however, ** it's application is brand new ** ... approach to be used during #llmtraining exercises.

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

A good introduction to measuring quality of LLMs and GenAI apps via a combination of automated & human evaluations: https://lnkd.in/gYe7scGS. This is now an industry standard and most companies use some version of this approach for design time testing/evaluation and monitoring for quality/trust in production. Share any interesting learnings/experiences you may have.

55

1 Comment

Really useful and simple explanation of in context learning and Retrieval Augmented Generation (RAG) that many see as a "silver bullet" to address LLM limitations. While RAG is undoubtedly super useful technique, it is doesn't fully solve LLM limitations, especially for domain-specific use cases prevalent in Enterprise LLM applications. Understanding advantages and limitations of various LLM-enhancing techniques such as RAG is critical to build AI products at scale. #largelanguagemodels #artificialintelligence #airevolution

31

Good operators of factories observe real outcomes in their actual context -- cf Toyota's "genchi genbutsu". 𝐆𝐨𝐨𝐝 𝐨𝐩𝐞𝐫𝐚𝐭𝐨𝐫𝐬 𝐨𝐟 𝐋𝐋𝐌 𝐚𝐩𝐩𝐥𝐢𝐜𝐚𝐭𝐢𝐨𝐧𝐬 𝐚𝐫𝐞 𝐧𝐨 𝐝𝐢𝐟𝐟𝐞𝐫𝐞𝐧𝐭. How and how often should you review LLM inputs and outputs? LLMs are dynamic and constantly evolving. Despite their impressive zero-shot capabilities and often delightful outputs, their failure modes can be highly unpredictable. For custom tasks, regularly reviewing data samples is essential to developing an intuitive understanding of how LLMs perform. ✅ In Summary, 𝐋𝐨𝐨𝐤 𝐚𝐭 𝐬𝐚𝐦𝐩𝐥𝐞𝐬 𝐨𝐟 𝐋𝐋𝐌 𝐢𝐧𝐩𝐮𝐭𝐬 𝐚𝐧𝐝 𝐨𝐮𝐭𝐩𝐮𝐭𝐬 𝐞𝐯𝐞𝐫𝐲 𝐝𝐚𝐲 Recent research highlighted that developers’ perceptions of what constitutes “good” and “bad” outputs shift as they interact with more data https://lnkd.in/gzus7Cqm

15

3 Comments

Another new paper from my group: this time examining the theoretical capabilities of transformers to apply classic graph algorithms, and then measuring them in practice.

20

3 Comments

New research tackles the fight against misinformation with AI-powered counter-narratives! A fascinating paper explores using Large Language Models (LLMs) to automatically evaluate the effectiveness of Counter Narratives (CNs) in combating online misinformation. The study reveals that LLMs, particularly chat-aligned models, excel at generating and evaluating CNs when fine-tuned with domain-specific data. This promising approach could revolutionize how we address misinformation and promote healthier online discourse. #misinformation #disinformation #cognitivesecurity #generativeAI #LLM #AI #techforgood https://lnkd.in/gHrVk9bs

11

Looking forward to discussing best practices for building enterprise AI/ML systems with Hira Dangol, Nitin Aggarwal, and Steven Eliuk at the AI Quality Conference (AIQCON) in San Francisco tomorrow. Event details below. #nvidia #AIQuality #ai #genaI #enterpriseAI

64

1 Comment