Tech leaders say open models are easier to customize, cheaper to run, and security risks are manageable.

So-called “unlearning” techniques are used to make a generative AI model forget specific and undesirable info it picked up from training data, like sensitive private data or copyrighted material. But ...

Traditionally, AI progress was constrained by one thing above all else: access to data. Not enough volume. Not enough ...

Nvidia's KV Cache Transform Coding (KVTC) compresses LLM key-value cache by 20x without model changes, cutting GPU memory costs and time-to-first-token by up to 8x for multi-turn AI applications.

A team of computer scientists at UC Riverside has developed a method to erase private and copyrighted data from artificial intelligence models—without needing access to the original training data.

Climate scientists are confronting a hard truth: some of the most widely used models are struggling to keep up with the pace and texture of real‑world warming. The physics at their core remains sound, ...

Is it possible for an AI to be trained just on data generated by another AI? It might sound like a harebrained idea. But it’s one that’s been around for quite some time — and as new, real data is ...

New benchmark study results show leading AI models, including ChatGPT, Claude, and Gemini, still lag humans in visual math reasoning.

To feed the endless appetite of generative artificial intelligence (gen AI) for data, researchers have in recent years increasingly tried to create "synthetic" data, which is similar to the ...

Researchers find large language models process diverse types of data, like different languages, audio inputs, images, etc., similarly to how humans reason about complex problems. Like humans, LLMs ...