New ‘Test-Time Training’ method lets AI keep learning without exploding inference costs

By allowing models to actively update their weights during inference, Test-Time Training (TTT) creates a "compressed memory" ...
NextBigFuture

Test Time Training Will Take LLM AI to the Next Level

MIT researchers achieved 61.9% on ARC tasks by updating model parameters during inference. Is this key to AGI? We might reach the 85% AGI doorstep by scaling and integrating it with COT (Chain of ...
ZDNet

AI models know when they're being tested - and change their behavior, research shows

Several frontier AI models show signs of scheming. Anti-scheming training reduced misbehavior in some models. Models know they're being tested, which complicates results. New joint safety testing from ...
VentureBeat

Hugging Face shows how test-time scaling helps small language models punch above their weight

In a new case study, Hugging Face researchers have demonstrated how small language models (SLMs) can be configured to outperform much larger models. Their findings show that a Llama 3 model with 3B ...
Wired

This Tool Probes Frontier AI Models for Lapses in Intelligence

Executives at artificial intelligence companies may like to tell us that AGI is almost here, but the latest models still need some additional tutoring to help them be as clever as they can. Scale AI, ...
InfoQ

Olmo 3 Release Provides Full Transparency into Model Development and Training

Unlock the full InfoQ experience by logging in! Stay updated with your favorite authors and topics, engage with content, and download exclusive resources. Vivek Yadav, an engineering manager from ...
TechCrunch

OpenAI partner says it had relatively little time to test the company’s o3 AI model

An organization OpenAI frequently partners with to probe the capabilities of its AI models and evaluate them for safety, Metr, suggests that it wasn’t given much time to test one of the company’s ...
SpaceNews

Space Force to revamp training for a new era of space conflict

Illustration of Earth orbit generated by AI using OpenAI’s DALL·E The U.S. Space Force is expanding its search for training and testing technologies and is now planning to put more than a half billion ...