All 3,244 AI stories, newest first · page 118 of 136
Researchers introduce AutomationBench, a new benchmark for evaluating AI agents on complex, cross-application workflows. It tests coordination, API discovery, and policy adherence in real-world business scenarios.
Researchers introduce ARES, a new framework to detect and mitigate systemic weaknesses in reinforcement learning from human feedback (RLHF). ARES addresses cases where both the reward model and the core LLM fail simultaneously, a critical vulnerability in current alignment methods.
Anthropic's powerful Mythos AI model, intended for cybersecurity, has been accessed by unauthorized users. The breach raises concerns about the misuse of advanced AI tools. The leak occurred through a third-party contractor, highlighting vulnerabilities in AI security protocols.
An AI-driven startup has successfully overturned thousands of denied health insurance claims. The company leverages machine learning to analyze claim denials and identify grounds for appeal, with notable support from Mark Cuban.
10x Science has secured $4.8 million to help researchers navigate the overwhelming number of AI-generated drug candidates. The startup aims to streamline pharmaceutical R&D by identifying the most promising molecules.
A curated list of small LLMs optimized for 16GB RAM devices, including Qwen3.5 9B and others. These models balance performance and efficiency for daily use.
Researchers introduced SynopticBench to evaluate vision-language models (VLMs) on weather forecasting tasks. The benchmark highlights the challenges of generating accurate meteorological discussions from chaotic atmospheric data.
Researchers introduce a novel framework that integrates large language models (LLMs) with Random Forests (RF) through reinforcement learning. This method enables iterative feedback between gradient-based and non-differentiable models, enhancing predictive performance.
QU-NLP's approach to the QIAS 2026 shared task leverages multi-stage QLoRA fine-tuning on Qwen3-4B for complex Islamic inheritance reasoning. This method enhances structured reasoning in legal domains with precise fractional calculations and rule-based decisions.
OpenAI’s updated image generator, ChatGPT Images 2.0, now pulls information from the web to create more sophisticated images. This new feature allows for more detailed and accurate AI-generated art based on single prompts.
OpenAI has launched GPT Image 2, a new AI model capable of generating high-quality images from text prompts. The model represents a significant leap in AI-driven image creation.
OpenAI has launched Codex Labs and partnered with major consulting firms to help enterprises integrate Codex into their software development workflows. The move comes as Codex reaches 4 million weekly active users.
A comprehensive survey explores domain-level data mixing techniques for optimizing large language model pretraining. The research underscores the importance of strategic data composition in improving efficiency and generalization under constrained resources.
Researchers tested four methods for identifying authors based on stylistic features in Japanese reviews, aiming to support future dark web actor analysis. The study highlights the potential of BERT-based approaches for authorship attribution.
A new arXiv paper critiques non-symbolic methods like SHAP for lacking rigor in high-stakes AI explanations. It advocates for symbolic approaches to improve trustworthiness. This could reshape the XAI landscape.
Researchers have introduced the first benchmark for extracting claims from multimodal social media posts, addressing a gap in automated fact-checking. The dataset includes text combined with images like memes and screenshots, challenging traditional methods.
NeoCognition, founded by an OSU researcher, has raised $40M to build AI agents capable of mastering any domain through human-like learning. The funding highlights the growing interest in advanced AI learning mechanisms.
Meta plans to use employee mouse and keyboard tracking data to train AI agents. This raises privacy concerns and highlights the ethical challenges of AI training methods.
Researchers introduced LiFT, a framework that improves large language models' ability to handle longitudinal NLP tasks. The method uses instruction fine-tuning to better track temporal changes in human behavior and opinions.
Hyatt has deployed OpenAI's ChatGPT Enterprise, featuring GPT-5.4 and Codex, to its entire global workforce. The move aims to enhance productivity, operations, and guest experiences across the hospitality chain.
Researchers propose HalluSAE, a novel method to detect hallucinations in LLMs by modeling them as critical shifts in latent dynamics. This approach addresses the dynamic nature of hallucinations often overlooked by existing methods.
Researchers introduce GoCoMA, a multimodal framework to identify the source of LLM-generated code. This addresses security and licensing concerns by analyzing code stylometry and binary image representations.
Researchers introduce GeoRepEval, a framework to evaluate large language models' robustness to different geometric problem representations. The study highlights that current benchmarks overlook representation invariance, masking potential failures.
Researchers introduce EchoChain, a new benchmark to evaluate AI assistants' ability to manage interruptions during conversations. The study identifies key failure patterns in current systems when users interrupt mid-response.