IEEE Spectrum on MSN

Why are large language models so terrible at video games?

AI models code simple games, but struggle to play them ...

MLCommons Releases New MLPerf Inference v6.0 Benchmark Results

Today, MLCommons ® announced new results for its industry-standard MLPerf ® Inference v6.0 benchmark suite. This release includes several important advances that ensure the benchmark suite tests ...
Morning Overview on MSN

Google’s TurboQuant claims 6x lower memory use for large AI models

Google researchers have proposed TurboQuant, a method for compressing the key-value caches that large language models rely on ...

Intel Arc Pro B70 Delivers Up to 80% AI Inference Boost in MLPerf v6.0 Benchmarks

Intel Arc Pro B70 delivers up to 80% faster AI inference in MLPerf v6.0 benchmarks, with strong GPU and CPU performance gains ...
VentureBeat

Beyond generic benchmarks: How Yourbench lets enterprises evaluate AI models against actual data

Every AI model release inevitably includes charts touting how it outperformed its competitors in this benchmark test or that evaluation matrix. However, these benchmarks often test for general ...
EurekAlert!

MathEval: a comprehensive benchmark for evaluating large language models on mathematical reasoning capabilities

This study introduces MathEval, a comprehensive benchmarking framework designed to systematically evaluate the mathematical reasoning capabilities of large language models (LLMs). Addressing key ...
VentureBeat

Researchers warn of 'catastrophic overtraining' in LLMs

A new academic study challenges a core assumption in developing large language models (LLMs), warning that more pre-training data may not always lead to better models. Researchers from some of the ...