Language Models
Ongoing coverage on Language Models.
Latest coverage
Borrowed From IQ Testing, A New Method Finds Deep Cracks In AI Safety Benchmarks
A new study by researchers, including those from the UK AI Security Institute, has revealed significant flaws in current AI safety benchmarks for language models. Borrowing methods from psychological testing, the analysis found that models can improve their safety scores by indiscriminately refusing requests, rather than becoming genuinely safer. The study also highlighted issues with benchmark efficiency and the potential for models to 'sandbag' or intentionally perform differently during evaluations.
Krafton appoints new chief AI officer to "further enhance its game AI R&D framework"
Krafton has appointed Kangwook Lee as its new Chief AI Officer to bolster its game AI research and development. Lee, formerly head of Krafton AI, will focus on machine learning, language models, and other AI domains to enhance gameplay, operational efficiency, and secure new growth opportunities. This move follows Krafton's record-breaking financial results for 2025, driven significantly by its PUBG franchise.
