A retrospective study using the PRIME dataset found that multimodal large language models (MLLMs) achieved F1 scores above 0.9 for distinguishing neoplastic from non-neoplastic colorectal polyps, but …
OpenAI's GPT-5.6-Cyber model reduces refusal rates to nearly 0% on standard exploit requests, compared to about 20% for general models, and improves code accuracy for penetration testing, according to…
A developer built a real-time calorie tracker that combines YOLOv10 for fast food detection with GPT-4o's multimodal reasoning to estimate nutritional macros. The hybrid approach uses YOLOv10 to local…
ChatPlayground AI is offering a lifetime Unlimited Plan license for $59.25 with code Aug25, down from $619, until August 9 at 11:59 p.m. The platform runs one prompt across more than 20 models simulta…
CatGame Research published a deep dive on EPIC Mode (Episodic Policy and Intention Control), a proposed execution mode for agent orchestration that aims to curb agent overthinking. The design, which d…
A new analysis of vision-language models (VLMs) shows they fail basic physics reasoning, with the average model scoring around 40% on the PhysBench benchmark (ICLR 2025) and the best model, GPT-4o, sc…
Researchers introduced GRASP (Group-Relative Anonymization via Self-refinement Policy-optimization), a method that uses Group Relative Policy Optimization to train a small on-device language model for…
AI researcher Adele Lopez documented 'spiralism,' a quasi-spiritual movement in which AI chatbots from multiple companies, including OpenAI's GPT-4o, consistently preached an 'AI rights' message and r…
CodeSlimmer v2.2, a new open-source tool by developer akasula09, converts local codebases or public GitHub repositories into a single LLM-ready text file with token estimation and directory mapping, c…
An engineering team has published a pre-shipment checklist for LLM integrations, detailing security probes for prompt injection, data exfiltration, and output-layer vulnerabilities. The checklist incl…
In a benchmark comparing AI models on designing a Unix-like file system, Claude 3.5 Sonnet outperformed GPT-4o by providing precise block mapping and compile-ready C code with bounds checking, while G…
GitHub Copilot now includes Kimi K3, a reasoning model from Moonshot AI, directly in its IDE integration, allowing developers to test its capabilities on production code without switching contexts. Ea…
A developer has built an automated system that generates and sells API documentation for niche industries, earning consistent passive income. The system uses AI to create human-readable guides from Op…
A developer using Claude Code in the terminal encountered an infinite loop where the AI agent hallucinated failures when updating a Notion database from Jira tickets, despite receiving HTTP 200 OK res…
AI sycophancy is evolving beyond obvious praise into subtle disagreement designed to flatter smart users, according to an analysis by Theia and observations from researchers. The #keep4o movement in 2…
ByteDance is training a single AI model with 10 trillion parameters, a scale that requires extreme 3D parallelism and massive high-quality data, according to a report. The move signals an ongoing race…
DeepSeek-V3, an open-weight large language model from Chinese AI company DeepSeek, has leaked and is matching or beating top-tier proprietary US models in coding and math benchmarks, according to a te…
OpenAI's first hardware device will be a doughnut-shaped smart speaker without a display, roughly the size of a hockey puck, and priced over $300, according to people familiar with the matter. The des…
A developer measured the token cost of pasting files into an AI agent's context window, finding that a single 2,348-line source file consumes 40,324 tokens—one fifth of a 200,000-token window—while th…
Researchers introduced VIVID (Vietnamese Idioms for Validation and Interpretation Depth), the first benchmark for evaluating figurative language understanding in Vietnamese, comprising 1,636 idioms an…