Page 27: Research news on Large language models

Large language models are high-capacity neural sequence models trained on massive text and multimodal corpora to perform language understanding, generation, and reasoning. Current work examines their internal representations, cognitive and social behavior analogies to humans, and limitations in mathematical, causal, and strategic reasoning. Research also addresses alignment with human values and brain activity, safety and security vulnerabilities, privacy and de-anonymization risks, cross-lingual and sociocultural biases, scaling and efficiency laws, and frameworks for tool use, multi-agent interaction, and domain-specific deployment.

Artificial Intelligence

Study finds AI can safely assist with some software annotation tasks

A dystopian future where advanced artificial intelligence (AI) systems replace human decision-making has long been a trope of science fiction. The malevolent computer HAL, which takes control of the spaceship in Stanley Kubrick's ...

Security

Five crucial ways LLMs can endanger your privacy

The privacy concerns around large language models like ChatGPT, Anthropic and Gemini are more serious than just the data the algorithms ingest, according to a Northeastern University computer science expert.

Artificial Intelligence

AI language models show bias against regional German dialects

Large language models such as GPT-5 and Llama systematically rate speakers of German dialects less favorably than those using Standard German. This is shown by a recent collaborative study between Johannes Gutenberg University ...

Artificial Intelligence

Mind readers: How large language models encode theory-of-mind

Imagine you're watching a movie, in which a character puts a chocolate bar in a box, closes the box and leaves the room. Another person, also in the room, moves the bar from a box to a desk drawer. You, as an observer, know ...

page 27 from 38