Koby Choy

My Blogs

  • Is Neurosymbolic AI the future Of Truth-Seeking AI? – part 1: LLM Meets Reasoning

    Is Neurosymbolic AI the future Of Truth-Seeking AI? – part 1: LLM Meets Reasoning

    Earlier this month, I had the privilege of engaging in a thought-provoking conversation with a senior IT professional about recent developments in Large Language Models (LLMs) and artificial intelligence in general. Our discussion centered on OpenAI’s new model, O1, and explored potential avenues for improving System 2 thinking in AI models, including enhanced comprehension and… Read more

  • Finetuning Phi3-mini-4k-instruct on custom Data + CPU Inference

    Finetuning Phi3-mini-4k-instruct on custom Data + CPU Inference

    Small Language Models (SLM) In the past two years, robotics has undeniably gained increasing attention due to exciting advancements in Large Language Models (LLMs). Many of our expectations for robots, such as engaging in natural conversations and taking actions based on spoken commands, can only be realized with the help of these LLMs. However, despite… Read more

  • Navigating the Challenges of Data Drift in Real-World AI Applications

    Navigating the Challenges of Data Drift in Real-World AI Applications

    Personal Story Two years ago, our team embarked on an ambitious project: developing an AI-powered receipt information extraction and validation system with no human intervention. The concept was straightforward: users would photograph their receipts, and our intelligent system would process these images through an intricate pipeline—transforming photos into actionable data using image processing, OCR engines,… Read more

  • Notes on the A.I. Act of the European Union (Prohibited AI Practises)

    Notes on the A.I. Act of the European Union (Prohibited AI Practises)

    The AI Act, which came into effect on August 1st, 2024, marks a significant step in regulating artificial intelligence (AI). This groundbreaking legislation establishes the first comprehensive legal framework for AI in Europe, aiming to ensure public trust in AI services. [1] The Act employs a risk-based approach, categorizing AI applications into four tiers: unacceptable… Read more

  • “Eyes beat Memory”

    “Eyes beat Memory”

    Earlier this year, I took a course on data visualization, and I have to be honest – I was blown away. Coming from a technical background, my friends and I focused more on data transformation, imputation, and forming hypotheses for statistical tests than on how to present our findings in slides or papers. We didn’t… Read more

  • Political Biases in Language Models (Part II – Evaluation of Politial Leanings)

    Political Biases in Language Models (Part II – Evaluation of Politial Leanings)

    In my previous discussion, I delved into the paper by Feng et al., published at ACL 2023, which explored the intricate relationship between political biases in large language models (LLMs) and their performance on downstream tasks like hate speech and misinformation detection. Now, let’s dive deeper into the solutions proposed by the researchers to address… Read more

Design a site like this with WordPress.com
Get started