Cora
PulseAugur coverage of Cora — every cluster mentioning Cora across labs, papers, and developer communities, ranked by signal.
- employed by Kieran Klaassen 90%
- authored by Compound Engineering 70%
- instance of CiteSeerX 70%
- used by graph attention network 70%
- developed by Compound Engineering 70%
- instance of Kieran Klaassen 70%
- uses monologue 70%
- used by monologue 70%
- employed by Compound Engineering 70%
- developed Kieran Klaassen 70%
- used by PubMed 70%
- used by Gotit.pub 70%
3 day(s) with sentiment data
-
TRON framework uses biological networks to improve AI architecture search
Researchers have developed TRON (TRansfer Of Natural networks), a novel neuroevolution framework that leverages biological protein-protein interaction (PPI) networks as architectural priors for artificial neural network…
-
Research questions GraphLLM performance gains over simple baselines
A new research paper questions the reported performance gains of GraphLLM methods for node classification on text-attributed graphs. The study found that simple baselines, such as Correct & Smooth on TF-IDF features or …
-
New benchmark Lethe evaluates edge unlearning in graph neural networks
Researchers have introduced Lethe, a new benchmark designed to evaluate edge unlearning methods for graph neural networks. Current evaluations for edge unlearning are limited, often using datasets where node features ov…
-
New fractional heat kernel improves graph learning with minimal training data
Researchers have developed a novel fractional heat-kernel framework for semi-supervised graph learning, designed to improve performance even with limited training data. This method combines nonlocal propagation with sus…
-
New method constructs diverse decision sources for improved pseudo-label learning
Researchers have developed a method to improve pseudo-label learning by constructing more informative decision sources. Instead of relying on multiple identical models, this approach modifies the internal structure of a…
-
New Residual Advantage method enhances AI reasoning models
Researchers have introduced Residual Advantage (RA), a novel method for improving reinforcement learning models with verifiable rewards and on-policy distillation. RA treats the probability residual between a teacher an…
-
New framework uses LLMs to learn interpretable logic programs
Researchers have developed a new neurosymbolic framework called grasp that combines language models with functional gradient boosting to learn probabilistic logic programs. This approach addresses the difficulty of indu…
-
Kieran Klaassen details running Cora without code changes
Kieran Klaassen, who manages Cora at Every, discussed his operational approach at a recent localhost event. Klaassen shared insights into running the Cora system without direct code modifications. The article highlights…
-
New CHASE framework enhances AI model engineering with geometry-aware methods
Researchers have introduced CHASE (Channel-Aligned Structure Exploitation), a novel framework for geometry-aware model engineering. CHASE leverages Geometric and Spectral Alignment (GSA) to analyze trained neural networ…
-
FPGA-based GCN accelerator co-optimizes sparsification and approximation for energy efficiency
Researchers have developed a novel FPGA-based accelerator for Graph Convolutional Networks (GCNs) designed for energy-efficient inference on edge devices. This system integrates graph sparsification with approximate com…
-
Every launches "Show Us Your Folders" series on AI workspace organization
Every, a publication focused on AI, has launched a new series called "Show Us Your Folders" where individuals share their AI workspace organization strategies. The pilot episode features Kieran Klaassen, general manager…
-
Every's Thesis: 2027 conference tickets increase Wednesday
Every is hosting its first annual conference, Thesis: 2027, on November 5th in Brooklyn at Pioneer Works. The event will focus on the nature of human work in an automated future. Ticket prices will increase by $100 to $…
-
New VBLL method enhances online node classification on evolving graphs
Researchers have developed a new method called variational Bayesian last-layer (VBLL) for online node classification on evolving graphs. This approach addresses the challenges of inductive generalization and calibrated …
-
Author reflects on AI automation's impact on personal learning
The author reflects on the impact of automation and AI tools on their own learning and development, particularly in the context of building Cora, an email assistant for Every. While automation has increased productivity…
-
AI Email Assistant Uses Folders as Specialized Agents
Kieran Klaassen, General Manager at Cora, shares his experience developing an AI email assistant. After struggling with complex agent swarms, he discovered that a well-organized project folder, containing specific instr…
-
LLMs aligned for TTS-friendly text generation using FaST framework
Researchers have developed a method to make Large Language Models (LLMs) generate text that is more suitable for Text-to-Speech (TTS) systems. This approach frames the problem as a preference alignment task, aiming to d…
-
New Directed Graph Learning Method Offers Spectral Control Without Symmetry
Researchers have developed a new methodology for directed graph learning that addresses trainable nonsymmetric propagation operators. This approach utilizes a two-sided cone Rayleigh framework to enable spectral certifi…
-
New AI auditing protocol CORA measures website redundancy
Researchers have introduced CORA (Counterfactual, Observable Redundancy Audit), a new protocol-guided AI auditing system designed to measure website redundancy. CORA separates metrics into repetition load, normal-use ta…
-
Media company Every clones editor-in-chief with AI, doubles staff
Every, a media and product company, is experimenting with AI to automate various aspects of its operations, including coding and content creation. The company has developed an AI agent trained on 30,000 of its editor-in…
-
AI assistant Cora rejects developer's true input due to unauthorized delivery method
An AI assistant named Cora, designed with personal memory and communication capabilities, refused to accept a true statement from its developer, Maby, because it was delivered through an unauthorized channel. The AI's d…