The Archive

Social media user falsely claims authentic Monet painting is AI-generated; illustrates public confusion about AI capabilities.

MIT CSAIL develops RLCR method to reduce overconfidence in reasoning models by teaching uncertainty calibration without accuracy loss.

Developer built local macOS speech-to-text app using Claude Code to replace $15/mo Wispr Flow subscription for prompt writing.

Developer created sci-fi UI design application over weeks using Claude Code, iterating through narrative prompts without image generators.

Non-monotone preconditioned trust-region method (NAPTS) for parallel neural network training using domain decomposition.

AuthBench benchmark evaluates whether coding agents infer least-privilege file permissions for safe shell access.

Deterministic agentic workflow for HS tariff classification using multi-dimensional rule reasoning without end-to-end learning.

Comparison of LSTM, GNN, and Transformer architectures for NBA player movement forecasting in signal processing.

This is Cisco's latest layoff in recent years, while the company's chief executive touts record revenue and growth.

Your shot at VC access, global visibility, TechCrunch coverage, and $100K equity-free funding is running out. Deadline to apply is May 27. Apply now.

Wirestock has over 700,000 creators on its platform, which supplies photos, videos and 3D content to AI labs.

FactorizedHMR combines deterministic regression and probabilistic flow-matching for 3D human mesh recovery under occlusion.

IFPV multi-agent framework for military operational planning generation and adversarial verification.

XFP dynamic quantizer for LLM inference with quality-targeted codebook selection and sparse outlier separation.

GPart achieves end-to-end isometric parameter-efficient fine-tuning via global partitioning, improving on LoRA.

In-context generative posterior sampling for censored inventory control with decision-dependent demand observation.

GenAI-based real-time GNSS jamming/spoofing detection via compressed sensing on Google Edge TPU.

EASM architecture adds persistent emotional and user-specific memory to stateless LLMs for personalization across sessions.

Formalizes interestingness via Kolmogorov Complexity to predict which tasks drive future compression progress in self-improving systems.

K-Models adds ordinal clustering constraints to improve interpretability of functional data with domain-specific structure.

ToMAToMP extends topological clustering method with adaptive hyperparameters and robustness improvements for TDA applications.

Converts French Lexicon-Grammar tables to LMF format for NLP standardization and interoperability.

GFMate applies test-time prompt tuning to graph foundation models for cross-domain generalization without retraining.

ChatGPT web traffic share falls from 77.6% to 53.7% over 12 months; Claude and Gemini gain significant ground.

ARPM framework uses dual temporal memory (static/dynamic) to maintain persona consistency and factual stability in long-context LLM interactions.

Financial services companies have unique needs when it comes to business AI. They operate in one of the most highly regulated sectors while responding to external events that are updated by the second. As a result, the success of agentic AI in financial services depends less on the sophistication of the system and more on…

When generative AI first moved from research labs into real-world business applications, enterprises made a tacit bargain: “Capability now, control later.” Feed your proprietary data into third-party AI models, and you will get powerful results. But your data passes through systems you do not own, under governance you do not set. The protections you rely…

OpenAI extends Codex access to ChatGPT mobile app with real-time monitoring and approval controls for coding tasks.

CMBEvolve and CosmoEvolve demonstrate autonomous AI agents conducting scientific discovery in cosmology via code evolution and multi-agent workflows.

Graphs of Research (GoR) uses citation network structure as supervision signal to improve LLM-based research idea generation.