Victor Kostyuk

I'm a computer scientist who now works in biotech and writes poetic blogposts in Russian. How come? What do AI, poetry, and biology have in common? Language. At school I hated Russian grammar class. Learning a language by its rules is like learning to dance from a book. Even if you master a thousand rules and a million exceptions, it doesn't guarantee you'll write well. A natural language isn't designed; it's evolved. It wasn't developed, it developed. Studying grammar helps, but the only real way to master a language is through developing intuition – by reading, writing, and speaking a lot. Turns our the same applies to computers. For years, we tried to teach them natural language through complex logic, grammar rules, and it went even worse than my Russian lessons. Remember those dumb chatbots from a decade ago? But then, large language models (AI) arrived, and instead of programming on linguistic rules, we just had them "read" a lot, basically every text in existence. And now, ChatGPT writes in almost every natural language better than humans. But the most exciting part lies ahead. See, biology is also a language in a way — the Language of Life. Amino acids are the "letters," proteins the "words," biochemical pathways the "sentences," and every cell is an "epic" of 94 million words. This language, too, is the product of evolution, yet far more complex than Russian. For centuries, we've tried unsuccessfully to decode it by rules, and we've only managed to create something akin to the first chatbots. But soon, AI will change everything. At OPTIC we're building a translator at the crossroads of AI and biology. I believe that once we find an optimal, scalable way to collect daya and let machines read all the existing "texts in the language of life," we'll develop a model capable of writing in it. Can you imagine the possibilities when the "ChatGPT moment" happens in biology? A cure for every disease? Personalized therapy for everyone? An end to aging? — "Ok, BIOPTIC, cure Parkinson's disease." — "HDEKKPKNFFGWKDHEQCMHPKSIM…🫶" This is what drives me to work every morning.

442

15 Comments

Wellfound asked us why a superstar chip designer should join ChipStack. To me, the answer is simple: - The world of designing and verifying chips is undergoing a transformative shift with generative AI. In 2025, we will see its fruition with mature tools. At ChipStack, you can be at the forefront of this revolution. - Anchor your existing skills in design and verification while diving into new territories like AI and enterprise software. Want to grow without leaving behind what you excel at? This is your chance. - If you’ve experienced the inefficiencies of traditional chip design and verification firsthand, why settle for the status quo? Be part of the change! We still have an open role for chip designers: https://lnkd.in/gGZJHNFf. Thanks Wellfound for being an awesome partner!

67

MarkovML was founded on the belief that your data is the ultimate competitive advantage. As Peter Norvig famously stated, "𝙈𝙤𝙧𝙚 𝙙𝙖𝙩𝙖 𝙗𝙚𝙖𝙩𝙨 𝙘𝙡𝙚𝙫𝙚𝙧 𝙖𝙡𝙜𝙤𝙧𝙞𝙩𝙝𝙢𝙨 𝙗𝙪𝙩 𝙗𝙚𝙩𝙩𝙚𝙧 𝙙𝙖𝙩𝙖 𝙗𝙚𝙖𝙩𝙨 𝙢𝙤𝙧𝙚 𝙙𝙖𝙩𝙖." This adage is particularly relevant in today's landscape, where Large Language Models (LLMs) are becoming commoditized. This has again been proven by the LLama3.1 paper (links in the comments). The difference between success and failure often hinges on having right data and how effectively it's managed and utilized by business. Here are some of the best practices to maximize value for your enteprise. 👉 𝗙𝗼𝗰𝘂𝘀 𝗼𝗻 𝗗𝗮𝘁𝗮 𝗤𝘂𝗮𝗹𝗶𝘁𝘆 ✨ Regular Audits: Ensure data accuracy and integrity through consistent and thorough audits. ✨Validation and Updates: Ensure quality metrics are in place to ensure consistency and relevancy. 👉 𝗨𝗻𝗱𝗲𝗿𝘀𝘁𝗮𝗻𝗱 𝗬𝗼𝘂𝗿 𝗗𝗮𝘁𝗮 𝗗𝗶𝘃𝗲𝗿𝘀𝗶𝘁𝘆 ✨ Inclusive Data Collection: Diversify your data inputs to ensure your datasets genuinely represent global perspectives and scenarios, enhancing the applicability and relevance of your models. ✨ Broad Representation: Strive for datasets that encapsulate a variety of demographics, industries, and use cases. This diversity helps improve the generalization of your models across different contexts and applications. 👉 𝗘𝗻𝘀𝘂𝗿𝗲 𝗗𝗮𝘁𝗮 𝗶𝘀 𝗨𝗻𝗯𝗶𝗮𝘀𝗲𝗱 ✨Bias Detection: Models are biased because the data used to train them is. Identify & fix the root cause. ✨Fairness Metrics: Establish fairness metrics and ensure your models deliver equitable outcomes across all user groups. 👉 𝗔𝗱𝗼𝗽𝘁 𝗮 𝗣𝘂𝗿𝗽𝗼𝘀𝗲-𝗗𝗿𝗶𝘃𝗲𝗻 𝗔𝗽𝗽𝗿𝗼𝗮𝗰𝗵 𝘁𝗼 𝗠𝗼𝗱𝗲𝗹𝗶𝗻𝗴 ✨Clear Objectives: Ensure you establish your expectations from the model and understand the trade-offs. ✨Alignment with Business Goals: Models scale efforts constrained by time and resources. Have clear business metrics and measure model performance against these to maximize value. 👉 𝗘𝗾𝘂𝗶𝗽 𝗬𝗼𝘂𝗿 𝗧𝗲𝗮𝗺𝘀 𝘄𝗶𝘁𝗵 𝘁𝗵𝗲 𝗥𝗶𝗴𝗵𝘁 𝗧𝗼𝗼𝗹𝘀 Give your knowledge workers the right tools to automate routine data tasks and efficiently transform raw data into strategic insights so the business can make informed decisions faster. Superior data and the right tools aren't just assets; they're your competitive edge—invest in them for long-term success. #DataQuality #DataDiversity #AI #MachineLearning #BusinessStrategy #MarkovML #GenerativeAI #GenAI #Llama3.1

31

3 Comments

Love this paper that tests correlation between humans and evaluation models (LLM as a judge). Few big takeaways 1. Cohen's Kappa score is a better measure of accuracy than pure percentage agreement because it takes into account the bias of random agreement/correlation. 2. Humans only agree 97% of the time! This shows how alignment isn't an all or nothing game but rather a distribution where the goal is to get as close as possible without being perfect cause then you risk overfitting against the human labeler who produce the dataset. 3. Open source models are the way to go for evaluation models - performance is comparable (ex Llama3-70B vs GPT-4) and there are so many more benefits from a transparency, explainability, and reliability perspective. Really great stuff from the team at UMass. If you guys have any other cool papers regarding the LLM-as-a-judge space, would love to hear more! Paper Link: https://lnkd.in/edT8P8Tr

85

Parsing tables from long pdfs is hard today, but getting easier everyday. Retrieval of mid sections of the document is hard today, but getting easier everyday. Taking a judgement call, getting everyone on the same page, decision making with imperfect information, communicating to inspire, asking the right questions, having an agenda is hard today, and will remain hard tomorrow. #Skills #AIAge #HumanEdge

17

6 Comments

One new interesting application of LLMs: CAIS has released a forecasting bot that provides probability estimates for any question you ask. For example, to the question "Will GPT5 be released before 2025?", it answers that it's 70% likely. The bot works by searching for relevant information online, feeding it to GPT-4o, asking a few guiding questions, and then requesting a probability estimate. What's impressive is that this bot outperforms experienced human forecasters, even though GPT-4o wasn't specifically trained for this task. This kind of application shows that AI could be of great help in improving decision-making in the future. Link to the blog post: https://lnkd.in/gjP7JQgz Link the bot if you want to try it yourself: https://forecast.safe.ai/

11

Wealth positively correlates with marriage rates Marriage positively correlates with having kids Yet somehow, wealth negatively correlates with fertility rates Seems impossible, right? Welcome to Simpson's Paradox, where aggregated data tells a story that completely falls apart when you look at the subgroups. What's really happening: Poor families who do have kids tend to have more of them, overwhelming the fact that a higher percentage of wealthy people become parents at all. The crude relationships mask the underlying mechanics. Berkeley's famous gender bias case showed departments that individually favored women applicants somehow discriminated against them in aggregate. Hospital A beats Hospital B in both simple and complex surgeries, yet has worse overall outcomes. The deeper point is that causality is fractal. Change your level of analysis and the entire story can flip. Makes you wonder how many of our cherished societal narratives are Simpson's paradoxes in disguise. tl;dr - The map is not the territory, and sometimes the map isn't even a map of the territory you think it is.

26

4 Comments

LLMs are terrible at basic math. Why? Here's why: Unlike people, LLMs don’t conceptualize mathematics. They simply recognize patterns in the training data (text) they’ve learned from. When asked what 2 + 2 is, an LLM recalls instances of seeing ‘2’, ‘+’, ‘2’ and so on, allowing it to approximate solutions but not match the precision of a calculator. That’s how we end up with LLMs asserting that 2 + 2 = 5. But for analytical teams, like finance, that makes LLMs really hard to adopt in day-to-day workflows. The productivity gains of having AI do calculations are negated by needing to fact-check and audit every calculation and number. LLMs are also not deterministic, so no amount of prompting is going to guarantee you’ll get the correct answer every time. And that’s just for basic math – the outputs I’ve seen from LLMs on more complex calculations and analyses can get very wonky. Add thousands of rows of data and you’re squarely in hallucination land. So what’s the solution? Calculate things manually? No. The good news is that this is changing quickly and there are options: 1 - AI is getting better at math. Researchers continue to train LLMs with more specialized datasets that include a much wider array of mathematical problems and solutions. The newest models, like OpenAI’s latest family of models, are incorporating things like symbolic reasoning and “chain of thought” methods to break complex problems into smaller, logical steps the LLM can more easily process. Results have been really promising and we’ve seen each new iteration of models continue to beat previous benchmarks. 2 - Leverage AI agents For intensive workflows, and to really make sure your outputs are correct, working with companies that are building high-performing AI agents is likely your best bet. Agentic workflows allow LLMs to be used for what they are great at (inferring user intent, making decisions, etc.) while outsourcing the things they struggle with (complex calculations) to other parts of a broader system. Text-to-SQL, “tool-use”, few-shot prompting, and other strategies can be combined with real-time data integrations to dramatically improve performance on analytical operations. At Concourse we’ve spent the better half of 2024 getting our AI Agents to be really good at math, and I’m so excited to continue growing our capabilities to tackle even more complex mathematical problems for our customers in 2025. Reach out if you’re interested in learning more about how we’ve built Concourse and the impact we’re already delivering for finance teams.

42

5 Comments

Attention Sports Enthusiasts - I'm excited to share that our recent work "MambaNet: A Hybrid Neural Network for Predicting the NBA Playoffs" was just published in SN Computer Science. Led by Reza K. (a PhD student in my group), the paper demonstrates how Feature Imitating Networks (a method co-invented by me and Dr. Tuka Alhanai from New York University) can be used to improve the prediction of NBA playoff outcomes. More specifically, MambaNet showed a 14% improvement the existing state-of-the-art approaches! This result is particularly interesting as (to date) Feature Imitating Networks have mostly been tested in healthcare contexts, where they've shown strong potential in a variety of applications (e.g. https://lnkd.in/g7ggj3sR). Please reach out if you'd like to learn more. https://lnkd.in/gy9r74DP

131

5 Comments

🔮RAG + KG: Powerful Combination for Translation Multi-AI Agents Large language models (LLMs) have shown great promise in machine translation, but they still struggle with contextually dependent terms, such as new or domain-specific words. This leads to inconsistencies and errors that are difficult to address. Existing solutions often depend on manual identification of such terms, which is impractical given the complexity and evolving nature of language. While RetrievalAugmented Generation (RAG) could provide some assistance, its application to translation is limited by issues such as hallucinations from information overload. =================== In this paper, researchers propose CRAT, a novel multi-agent translation framework that: 》 leverages RAG and causality enhanced self-reflection to address these challenges. This framework consists of several specialized agents: ✱ the Unknown Terms Identification agent detects unknown terms within the context, ✱ the Knowledge Graph (KG) Constructor agent extracts relevant internal knowledge about these terms and retrieves bilingual information from external sources, ✱ the Causality enhanced Judge agent validates the accuracy of the information, ✱ Translator agent incorporates the refined information into the final output. =================== This automated process allows for more precise and consistent handling of key terms during translation. Their results show that CRAT significantly improves translation accuracy, particularly in handling context-sensitive terms and emerging vocabulary. --- If your company is looking for an AI consultant to accelerate business growth, or if you are looking to build your career in AI, send me a message and I can help

4

In Marketing Mix Modeling #mmm, confounders like seasonality, holidays, or macroeconomic indices can distort the relationship between Ad Spend and Sales. Without adjustment, we might overestimate advertising’s impact. Accounting for confounders “unrotates” the data, revealing the true causal effect—a small step with big implications for accurate insights. This is why I’m creating automatic data import as part of MMM Labs platform. We know this data is crucial to the success of an MMM project, so it should be in the model automatically. Prof. dr. Koen Pauwels, Ernest Baver, Vishal Sharma, Mehmet Ozgun Kurt, Dr. Robert Kübler

30

2 Comments

I just published a new article on Seeking Alpha. Using Python, I analyzed 10 years of data for Enbridge's common and preferred shares. Python is an incredible tool! Here is the article. Enbridge: Common Shares Downgraded As Preferreds Offer Better Value https://lnkd.in/gh94cHGb

10

This is more of a "talking to myself in public" thing than any kind of fully fleshed out hypothesis, so take this with the appropriate grain(s) of salt. But if I were interested in AI and were looking for an angle to explore that might yield something interesting, I'd start working on learning "Multiscale Mathematics" (also known as "Multiscale Modeling"). This is rooted in my personal belief that there has to be a "multi-scale" aspect to intelligence because there is such an aspect inherent in our underlying physical reality.

2

This is the final part of the 3 part series of posts that shows how to Evaluate your LLM based RAG application. In the last part what I would like to show us how to Evaluate the Answer (Actual Response) faired against the Contexts Provided after Retrieving it from a Vector Database. The only metric that is available for this is Faithfulness. Where the Claims made in the Answer (Actual Response from the LLM) are checked against the Claims present in the Contexts Provided to the LLM to answer the user query. The way the Claims are obtained is by asking the same LLM or another LLM to present all the Claims made in the Answer, and then get all the Claims made in the Contexts that was provided to the actual LLM to generate the Answer. Then calculate the ratio between the two. #llm #evaluation #metric #ai

4

IMVC 2024 just released Dr. Amir Alush great talk about going from messy visual data to high quality datasets: https://lnkd.in/dGBDUUHQ . Try it yourself on your own data at https://lnkd.in/ds4vwrpP!

24

ok - given twitter/x is the place where bleeding edge R&D happens (if you can evade the AI hype and crypto posts successfully), I'm trying to figure out how much cross-over there is over here in what i think is more of the real world. 1. have you heard of the 'anon' entropix project? (uses parallel sampling, with a draft model, and a bunch of other inference-level 'hacks' to the tokenizer and decoder) 2. have you tested it? (i did for ~1 hour yesterday, not sure what to think yet) 3. has anyone seen anything neat with speculative decoding + draft models (haven't seen a lot in the open since medusa)? one thing i know i like about entropix is that it uses a base model (without chat/instruct tuning) + an instruct in parallel. base models are powerful - even 1B parameter ones. tldr: if there's any activity/comments on this post, i will be shocked, but genuinely curious to get the real world's perspective. if you can get really fantastic results by running two 1B parameter LLMs on certain tasks, that clears the way for running almost everything locally, on-device, privately, and only offloading to cloud/api services what you can't do locally.

12

7 Comments

Ever found yourself torn between Langchain and LlamaIndex, unsure which GenAI framework to use? What if you didn't have to choose? 🤔 I just wrote a blog post where I show how to combine the best of both frameworks to build an Agentic RAG System. This can be a great starting point for someone wanting to implement Agents for their GenAI use cases. 🤖 You can check it out here: https://lnkd.in/g8fqq5GG

15

1 Comment

Getting computers to talk to each other is hard. 🗣️ There's a lot of hype about LLMs eliminating structured data altogether, but that won't happen because, fundamentally, natural language lacks *precision*. Here's why I think data standards aren't going anywhere: 1) Precision 🎯 - Natural language is inherently ambiguous (unless you're speaking Esperanto). We need data standards for the same reason we use mathematical notation: to remove ambiguity. When someone writes "5/10," we need explicit context to know if that's a date, rating, or fraction. 2) Monologue vs. Dialogue 💬 - When humans book international flights, we naturally engage in back-and-forth: "Is that departure time in local or destination timezone?" With wire transfers, we clarify: "Is that account number IBAN or SWIFT format?" But machines typically communicate in one shot: request/response. Even with LLMs, network contraints reinforce this pattern. 3) Validation ✅ - Because there's no dialogue, machines need to get it right the first time. Data standards enable this through validation: Is this price in dollars or euros? Is this product identifier UPC or EAN? While this can get pretty wonky, it determines whether a transaction succeeds or fails. The evidence is already here. Even ChatGPT uses function calling to make its outputs more deterministic (https://lnkd.in/gSJBreWE). And interfacing with LLMs themselves requires structured protocols, as shown by initiatives like the Model Context Protocol (https://lnkd.in/g9ZQJ_rg). The future looks less like LLMs replacing standards, and more like LLMs-as-compilers: translating human input into precise, machine-readable formats. Thoughts? When have you seen ambiguous data cause real business problems? #AI #FHIR #Integration

26

3 Comments

Meta just dropped Llama 3 (70b) last week. Here's what you need to know. Trained on 15 trillion tokens of data. 😯 Llama 3 is leading the space in capabilities for math and reasoning. Scoring an 85 on the MMLU benchmark. (Massive Multitask Language Understanding) Compared to Claude 3's impressive 86.9 score. However, with 25 trillion less training tokens. Llama 3 definitely has the potential to exceed Claude 3. The most exciting development comes in the form of Meta's approach to AI. 𝗢𝗽𝗲𝗻 𝗦𝗼𝘂𝗿𝗰𝗲. Llama 3's being released to developers, Along with a "Getting Started" guide, as well as new trust/safety features. - Llama Guard 2 A beefier version of their Llama guard content classification system. - Code Shield A screener for potentially malicious, or improper code outputs. - CyberSec Eval 2 A system for protection against prompt injections and code interpreter abuse. 💡 Benefits of Open Sourcing Llama 3 Llama 3 is comparable to GPT-4 in terms of capabilities. Tech hardware companies like Nvidia and AMD are supporting the model, As well as Cloud services like AWS, and Hugging Face. By it becoming publicly available to whomever wants to design with it. ✅We'll quickly have many GPT-4 models entering the market. The competition is on! - I'm really excited to announce Hank and I's collaboration with Christoph Rosenboom Keep an eye out next Monday! #llama3 #AI #LargeLanguageModel

17

19 Comments