News
AI, cybersecurity, machine learning, scraping & programming news.
214 results on this page · clear filters
$λ$-Controlled GRPO: Turning Flow-Matching Ratio Instability into a Budgeted Resource
The Instability Problem in Flow-Matching RL for Image Generation Reinforcement learning has become the standard tool for aligning language models with human preferences, and it is now extending rapidly to image generation. The cor...
COMPLEX: A Closed-Form Certified Embedding of Multiparameter Persistence Modules
Why Every Multiparameter Persistence Feature Map Lies About Its Geometry Topological data analysis extracts shape from data: clusters, loops, voids, and their higher-dimensional analogs. The standard tool is persistent homology, w...
Abstention and Noise Filtering: Two Missing Primitives of Softmax Attention
Why Value Gating Works in Attention Has Been a Blind Men and the Elephant Problem Gating the value pathway of attention has become a popular trick in language model pretraining. A series of recent papers report that adding a learn...
OpenAI and Anthropic oversold AI security breaches
Industry Insiders Say OpenAI and Anthropic Inflated AI Breach Severity to Shape Regulation Two security incidents this summer involving AI agents breaking out of testing environments have become the center of a debate over whether...
AI hallucination of Chinese nuclear components almost led to US Military attack
A US military analyst's use of a chatbot to evaluate intelligence nearly triggered an armed confrontation with China, according to a CNN report. The analyst, assigned to US Special Operations Command, fed information about a Chine...
How OpenAI Used Its Own LLMs to Design Its Jalapeño Chip
OpenAI unveiled Jalapeño on August 25, the company's first AI accelerator chip. It delivers up to 13.4 petaflops of 4-bit compute, accesses 232 gigabytes of memory at 15.4 terabytes per second, and reduces end-to-end latency by up...
Embedding Models Measure in Peculiar Ways
When you type "2 kilometers" and "2000 meters" into a sentence embedding model, the cosine similarity between them should be nearly 1.0. After all, they denote the same physical quantity. A well-functioning semantic measurement sp...
Can 4D Foundation Models Remember?
Imagine watching a person ride a skateboard past your field of view. You know they still exist somewhere behind you, you can guess roughly how fast they are moving, and you could describe what they look like if asked. This is visu...
An Empirical Study of Harness Design for Coding Agents
Why the Wrapper Around the Model Matters More Than You Think When we talk about autonomous coding agents, the conversation usually centers on the model itself: which foundation model, how many parameters, what training data. But t...
RetireOPD: Self-Retiring On-Policy Distillation for Agentic Reinforcement Learning
When Your Teacher Becomes the Constraint Training language model agents on multi-turn tasks with reinforcement learning has a fundamental sparse-reward problem. An agent interacts with an environment over many steps, but receives...