Yuriy Semchyshyn

Compressing LLMs but worried about accuracy? 🎯 EoRA uses eigenspace low-rank approximation to compensate for errors—no retraining needed. A promising direction for scalable, task-adaptive LLMs. New paper from #NVIDIAResearch.

1

Compressing LLMs but worried about accuracy? 🎯 EoRA uses eigenspace low-rank approximation to compensate for errors—no retraining needed. A promising direction for scalable, task-adaptive LLMs. New paper from #NVIDIAResearch.

1

🧠💬 Memory in LLMs A practical guide showing how to implement conversational memory in LLMs using LangGraph, demonstrated through a therapy chatbot. Features code examples for basic retention, trimming, and summarization approaches. Learn to build memory-aware apps 👉 https://lnkd.in/gybcrV5v

967

21 Comments

Today’s guide by Reza Bahmanzadeh, AI Engineer on our Professional Services Team, teaches the essentials of serving large models with vLLM: https://lnkd.in/eHK9nwZm It builds from foundational neural network concepts, like transformers and attention, to introduce practical inference workflows, explore useful vLLM features and offer hands-on, operational guidance for efficient deployments. #LLMs #Serving #Inference #TechGuides

35

1 Comment

NVIDIA, Sheba Medical Center & Mount Sinai are joining forces to decode the 98% of the human genome yet unexplored using AI-powered genomic foundation models, advancing disease prevention & precision medicine worldwide.

2

In this post, we demonstrate how to integrate Amazon SageMaker HyperPod with Anyscale platform to address critical infrastructure challenges in building and deploying large-scale AI models. The combined solution provides robust infrastructure for...

1

Rachit Jain's walkthrough shows how to profile kernels with PyTorch tools, import traces into NVIDIA Nsight or Intel Advisor, and read the slanted ceilings that separate compute-bound from memory-bound regions.

5

NVIDIA TensorRT LLM provides a high-level Python LLM API, with its PyTorch-native architecture enabling developers to experiment with the runtime or extend functionality. Learn how these latest TensorRT-LLM optimizations boost reasoning inference performance in our recent blog post. 🖇️ https://lnkd.in/gx52zwTA #PyTorch #OpenSourceAI #AI #Inference #Innovation

94

2 Comments

Comparing models.. Finding that a small 1B #llm hallucinates less than a 4B for semantic to structural matching. Small llms for targeted tasks for the win. The new Meta llama3.2 1B is better than expected and cheap on token cost.

3

We just published our first technical research note. "Low-Latency Event-Driven Architecture for Real-Time Options Execution on IBKR" It covers how we eliminated compounding feed lag at SPY market open — from a growing 45-second backlog to stable sub-second bar sealing — through: → Queue-first callback offloading (Python GIL problem) → Per-venue DOM isolation across ARCA, ISLAND, IEX → Event-time binning across 6 independent intervals → Single source of truth for all technical indicators No signals. No alpha. Pure infrastructure. This is the kind of work that happens before a single trade fires — and the kind that determines whether your system survives the opening bell. 📄 Download: https://lnkd.in/g-XJsYwG #AlgoTrading #OptionsTrading #MarketMicrostructure #SystematicTrading #Python #QuantitativeFinance #TradingInfrastructure

1

With a heavy heart as we mourn the loss of many of our fellow Iranians, I’d like to share a brief professional update. Our paper “GenForge: A Multi-population Genetic Programming framework with Semantic-Preserving Feature Partitioning for classification and regression tasks” has been published in Software Impacts. 🔹 Paper (DOI): https://lnkd.in/ggfdySDb 🔹 Code (archived version used in the paper): https://lnkd.in/gUyhyDZD 🔹 Main repository + documentation/manual: https://lnkd.in/guhSeiZ9 This work was carried out as part of my PhD research at UTS. I’m grateful to my supervisor Prof. Amir H Gandomi, and to my co-authors Navid Yazdanjue, Hassan Gharoun, A/Prof. Mohammad Reza Nikoo, and Prof. Fang Chen, for their contributions and support. GenForge is an open-source Python package for interpretable symbolic modeling via multi-population Genetic Programming (GP). It integrates Semantic-Preserving Feature Partitioning (SPFP) to build complementary, low-redundancy multi-view partitions from single-source data, and then evolves multiple GP populations (one per view) whose outputs can be fused through ensemble mechanisms for robust performance while keeping the final model transparent and parsimonious. 🔍 Key highlights: ✅ Unified symbolic learning for both regression and classification (gpregressor, gpclassifier) ✅ SPFPPartitioner for multi-view partitioning that supports view-driven multi-population evolution ✅ A consistent workflow (initialize → evolve → report → predict) with reproducible example scripts ✅ Built-in diagnostics and reporting utilities (including HTML summaries) to support explainable modeling and reproducible research 📌 Why does this matter? Many real-world problems need models that are both accurate and explainable. GenForge is designed to make symbolic, interpretable modeling practical in modern workflows by combining multi-view partitioning with cooperative multi-population GP, enabling robust learning while preserving interpretability. #GeneticProgramming #SymbolicRegression #ExplainableAI #MultiViewLearning #EnsembleLearning #Python #OpenSource #ReproducibleResearch #SoftwareImpacts #PhDResearch #UTS

44

6 Comments

AI researchers are fascinated by Large Language Models. But Large Tabular Models? Nope, they don't get anything like as much research love. Our PhD student Xiangjian Jiang wants to change that. He has just been awarded a prestigious 2025 Global Google PhD Fellowship and hopes the accolade will help him highlight an under-explored area of AI research that has applications in realms from finance to health. https://lnkd.in/gryW5wVc

60

Is it likely that that in the future, that America will realize that the attack on Iran was done for the wrong reasons, and will regret the attack? AI Overview Based on developments and public opinion polling from early 2026, it is highly plausible that significant portions of the American public and future policymakers could come to regret the military actions taken against Iran, as immediate public and international reactions show substantial division and disapproval. Public and Political Disapproval: As of March 2026, polling indicated that a majority of Americans (roughly 56%–60%) disapproved of the U.S. decision to take military action in Iran. Opposition to Escalation: While strong support existed among specific political groups, a widespread concern about the lack of a clear strategy and the risk of a "pointless, expensive, blood-soaked war" exists. International Concern: International partners, including Canada, have expressed "regret" over the U.S. actions, characterizing them as a failure of the international order and suggesting they are inconsistent with international law. Potential for Retaliation and Instability: Concerns have been raised that the strikes, which reportedly targeted Iranian leadership, could lead to severe retaliatory attacks on U.S. soil. The conflict has resulted in significant casualties, including civilian deaths (such as reports of a school being struck), which can create long-term regret and reputational damage. Divided Views on Rationale: The conflict has been described by some officials and experts as lacking a convincing rationale, setting a dangerous precedent, and being driven by political lobbying rather than urgent national security necessity. If the conflict results in prolonged instability, high U.S. casualties, or limited success in achieving its stated goals, it is likely that, similar to other foreign interventions, a consensus could emerge that the action was taken for flawed or incomplete reasons.

TIL that Iran launched Ballistic missiles at The Weizmann Institute of Science in June 2025, destroying labs and years of work. I don't know how you justify targeting Research Campuses. I was looking for the best places to conduct SOTA research in computer science. I'm waiting on my Master's until these wars subside, but this is so not cool. No student or researcher should have to factor "will my university get bombed?" into their career planning. That should be a basic global consensus. For those that don't know, The Weizmann Institute of Science offers top notch MSc. Programs free of cost so that students can completely devote their time to Scientific Research and Study. I was getting excited right before this news infuriated me. To the students and researchers at Weizmann: I’m deeply sorry for your loss. Years of work should not be erased by a missile salvo. The least the rest of us can do is not look away, and keep insisting that universities must be off‑limits everywhere.

1

Redis 8 is GA and packed with new features, including new data types, faster commands, higher throughput, and faster replication. Learn more about our most performant and scalable version yet: https://lnkd.in/eFU5qPDs

23

1 Comment

Check out this new paper from my Microsoft colleague Perry Abdulkadir! Turning Responsible AI itself into a scientific and engineering practice is the key to ensuring the field has lasting impact.

9

2 Comments

A recent piece in MIT Technology Review highlights a major shift in the AI landscape: Chinese open-source models are rapidly improving, and in some cases rivaling Western systems. For startups, this isn’t just geopolitics. It’s about cost, plain and simple. Open-weight models can be self-hosted, fine-tuned, and optimized without recurring API fees. That reduces dependency on closed providers and creates room for margin, especially for AI-native startups where inference costs directly impacts the bottom line. While U.S. frontier labs like OpenAI and Anthropic continue pushing the performance frontier, their models often sit behind APIs with usage-based pricing. For many early-stage companies, that means higher variable costs, vendor concentration risk, and less architectural control. At the same time, the race for strategic government partnerships, including recent competition between OpenAI and Anthropic for contracts with the United States Department of Defense, shows how central AI infrastructure has become to a national security strategy. The frontier labs may be optimizing for safety, scale, and federal compliance but startups are optimizing for runway. History favors the player who drives costs to zero and China may be the rapidly gaining dark horse turning AI into a commodity before anyone else sees it coming. https://lnkd.in/g_VZdgsm

2