Papers.

Adaptive Recursive Query Optimization

Performance-critical industrial applications, including large-scale program, network, and distributed system analyses, are increasingly reliant on recursive queries for data analysis. Yet traditional relational algebra-based query optimization techniques do not scale well to recursive query processing due to the iterative nature of query evaluation, where relation cardinalities can change unpredictably during the course of a single query execution. To avoid error-prone cardinality estimation, adaptive query processing techniques use runtime information to inform query optimization, but these systems are not optimized for the specific needs of recursive query processing. In this paper, we introduce Adaptive Metaprogramming, an innovative technique that shifts recursive query optimization and code generation from compile-time to runtime using principled metaprogramming, enabling dynamic optimization and re-optimization before and after query execution has begun. We present a custom join-ordering optimization applicable at multiple stages during query compilation and execution. Through Carac, a custom Datalog engine, we evaluate the optimization potential of Adaptive Metaprogramming and show unoptimized recursive query execution time can be improved by three orders of magnitude and hand-optimized queries by 6x.

Levels of AGI for Operationalizing Progress on the Path to AGI

We propose a framework for classifying the capabilities and behavior of Artificial General Intelligence (AGI) models and their precursors. This framework introduces levels of AGI performance, generality, and autonomy, providing a common language to compare models, assess risks, and measure progress along the path to AGI. To develop our framework, we analyze existing definitions of AGI, and distill six principles that a useful ontology for AGI should satisfy. With these principles in mind, we propose "Levels of AGI" based on depth (performance) and breadth (generality) of capabilities, and reflect on how current systems fit into this ontology. We discuss the challenging requirements for future benchmarks that quantify the behavior and capabilities of AGI models against these levels. Finally, we discuss how these levels of AGI interact with deployment considerations such as autonomy and risk, and emphasize the importance of carefully selecting Human-AI Interaction paradigms for responsible and safe deployment of highly capable AI systems.

What We Learned from a Year of Building with LLMs

Vector Databases: What's Really New and What's Next?

Transformer BERT and GPT3

Sophia: A Scalable Stochastic Second-order Optimizer for Language Model Pre-training.

Scalable OLTP in the Cloud: What's the BIG DEAL?

Examines what actually has to change to make OLTP scale in cloud environments, separating the genuinely hard problems from the ones that cloud storage and elasticity have already solved.

Red Hat build of Apache Camel for Spring Boot on OpenShift for IBM Power

ReAcTable: Enhancing ReAct for Table Question Answering

Python 3 Data Visualization Using Google Gemini

Python 3 and Machine Learning Using ChatGPT / GPT-4

One Step of Gradient Descent is Provably the Optimal In-Context Learner with One Layer of Linear Self-Attention.

Linguistic Calibration of Long-Form Generations.

Large Language Models for Developers

Large Language Models as Tool Makers.

Large Language Models

Knowledge Graphs for RAG

High Fidelity Web Archiving of News Sites and New Media with Browsertrix

Google Gemini for Python

Detecting Hallucination and Coverage Errors in Retrieval Augmented Generation for Controversial Topics.

Data Literacy with Python

CSS3 and SVG with Perplexity

CSS3 and SVG with Meta AI

CSS3 and SVG with GPT-4

CSS3 and SVG with Gemini

CSS3 and SVG with Claude 3

Chain of Thought Empowers Transformers to Solve Inherently Serial Problems.

Building Agentic Systems in an Era of Large Language Models

BioNeMo Framework: a modular, high-performance library for AI model development in drug discovery

Beginning Python 3 with Grok-2

Apache Iceberg: The Definitive Guide - Data Lakehouse Functionality, Performance, and Scalability on the Data Lake

Reference on the Apache Iceberg open table format: table architecture and metadata layout, catalogs, engine integrations (Spark, Flink, Dremio, Trino), partitioning and schema evolution, table optimization and maintenance, streaming ingestion, and lakehouse governance and migration patterns.

Augmented Intelligence for Clinical Discovery in Hypertensive Disorders of Pregnancy Using Outlier Analysis

Objectives Clinical discoveries are heralded by observing unique and unusual clinical cases. The effort of identifying such cases rests on the shoulders of busy clinicians. We assess the feasibility and applicability of an augmented intelligence framework to accelerate the rate of clinical discovery in preeclampsia and hypertensive disorders of pregnancy-an area that has seen little change in its clinical management. Methods We conducted a retrospective exploratory outlier analysis of participants enrolled in the folic acid clinical trial (FACT, N=2,301) and the Ottawa and Kingston birth cohort (OaK, N=8,085). We applied two outlier analysis methods: extreme misclassification contextual outlier and isolation forest point outlier. The extreme misclassification contextual outlier is based on a random forest predictive model for the outcome of preeclampsia in FACT and hypertensive disorder of pregnancy in OaK. We defined outliers in the extreme misclassification approach as mislabelled observations with a confidence level of more than 90%. Within the isolation forest approach, we defined outliers as observations with an average path length z score less or equal to -3, or more or equal to 3. Content experts reviewed the identified outliers and determined if they represented a potential novelty that could conceivably lead to a clinical discovery. Results In the FACT study, we identified 19 outliers using the isolation forest algorithm and 13 outliers using the random forest extreme misclassification approach. We determined that three (15.8%) and 10 (76.9%) were potential novelties, respectively. Out of 8,085 participants in the OaK study, we identified 172 outliers using the isolation forest algorithm and 98 outliers using the random forest extreme misclassification approach; four (2.3%) and 32 (32.7%), respectively, were potential novelties. Overall, the outlier analysis part of the augmented intelligence framework identified a total of 302 outliers. These were subseque

A Primer on Gene Editing: What Does It Mean for Pathologists?

The value of mathematical modelling approaches in epidemiology for public health decision making

Wikipedia Multilingual Vandalism Detection Dataset

Wikipedia Multilingual Vandalism Detection Dataset

Uncovering Bias in Personal Informatics

Judging LLM-as-a-judge with MT-Bench and Chatbot Arena

Evaluating large language model (LLM) based chat assistants is challenging due to their broad capabilities and the inadequacy of existing benchmarks in measuring human preferences. To address this, we explore using strong LLMs as judges to evaluate these models on more open-ended questions. We examine the usage and limitations of LLM-as-a-judge, including position, verbosity, and self-enhancement biases, as well as limited reasoning ability, and propose solutions to mitigate some of them. We then verify the agreement between LLM judges and human preferences by introducing two benchmarks: MT-bench, a multi-turn question set; and Chatbot Arena, a crowdsourced battle platform. Our results reveal that strong LLM judges like GPT-4 can match both controlled and crowdsourced human preferences well, achieving over 80% agreement, the same level of agreement between humans. Hence, LLM-as-a-judge is a scalable and explainable way to approximate human preferences, which are otherwise very expensive to obtain. Additionally, we show our benchmark and traditional benchmarks complement each other by evaluating several variants of LLaMA and Vicuna. The MT-bench questions, 3K expert votes, and 30K conversations with human preferences are publicly available at https://github.com/lm-sys/FastChat/tree/main/fastchat/llm_judge.

A Data Analysis Pipeline for Automating Apple Trait Analysis and Prediction.

Using Learned Indexes to Improve Time Series Indexing Performance on Embedded Sensor Devices.

Exploratory Research on Global Warming and Weather Data Simulation for Increased Sustainability or Resilience to Hazards Induced by a Changing Climate

First M87 Event Horizon Telescope Results. IX. Detection of Near-horizon Circular Polarization

Polarimetric Geometric Modeling for mm-VLBI Observations of Black Holes

A Search for Pulsars around Sgr A* in the First Event Horizon Telescope Data Set

Comparison of Polarized Radiative Transfer Codes Used by the EHT Collaboration

The Event Horizon Telescope Image of the Quasar NRAO 530

FairXAI - A Taxonomy and Framework for Fairness and Explainability Synergy in Machine Learning

FairXAI - A Taxonomy and Framework for Fairness and Explainability Synergy in Machine Learning

TEE: Real-Time Purchase Prediction Using Time Extended Embeddings for Representing Customer Behavior

Accretion Flow Morphology in Numerical Simulations of Black Holes from the ngEHT Model Library: The Impact of Radiation Physics