X

Vous n'êtes pas connecté

Maroc Maroc - TECHXPLORE.COM - RSS news feed - 21/Feb 18:19

Why GPT can't think like us

Artificial Intelligence (AI), particularly large language models like GPT-4, has shown impressive performance on reasoning tasks. But does AI truly understand abstract concepts, or is it just mimicking patterns? A new study from the University of Amsterdam and the Santa Fe Institute reveals that while GPT models perform well on some analogy tasks, they fall short when the problems are altered, highlighting key weaknesses in AI's reasoning capabilities. The work is published in Transactions on Machine Learning Research.

Articles similaires

Sorry! Image not available at this time

Like human brains, large language models reason about diverse data in a general way

techxplore.com - 19/Feb 13:49

While early language models could only process text, contemporary large language models now perform highly diverse tasks on different types of data....

Sorry! Image not available at this time

DarkMind A Novel Backdoor Attack Exploits Reasoning Capabilities of Customized LLMs

itsecuritynews.info - 18/Feb 12:11

A groundbreaking study by researchers Zhen Guo and Reza Tourani at Saint Louis University has exposed a novel vulnerability in customized large...

LLMs Are Not Reasoning—They’re Just Really Good at Planning

unite.ai - 19/Feb 17:02

Large language models (LLMs) like OpenAI’s o3, Google’s Gemini 2.0, and DeepSeek’s R1 have shown remarkable progress in tackling complex...

Reinforcement Learning Meets Chain-of-Thought: Transforming LLMs into Autonomous Reasoning Agents

unite.ai - 02:42

Large Language Models (LLMs) have significantly advanced natural language processing (NLP), excelling at text generation, translation, and...

Making AI work for finance operations

newsday.co.tt - 20/Feb 05:44

Gavin Hinks CFOs have been urged to take care with how they attempt to apply artificial intelligence (AI) in their finance operations as research...

Sorry! Image not available at this time

Neuro-inspired AI framework uses reverse-order learning to enhance code generation

techxplore.com - 18/Feb 13:50

Large language models (LLMs), such as the model behind OpenAI's popular platform ChatGPT, have been found to successfully tackle a wide range of...

The Many Faces of Reinforcement Learning: Shaping Large Language Models

unite.ai - 13/Feb 14:41

In recent years, Large Language Models (LLMs) have significantly redefined the field of artificial intelligence (AI), enabling machines to understand...

Top AI Models are Getting Lost in Long Documents

unite.ai - 13/Feb 14:43

A new study from researchers at LMU Munich, the Munich Center for Machine Learning, and Adobe Research has exposed a weakness in AI language models:...

Transformers and Beyond: Rethinking AI Architectures for Specialized Tasks

unite.ai - 08/Feb 20:12

In 2017, a significant change reshaped Artificial Intelligence (AI). A paper titled Attention Is All You Need introduced transformers. Initially...

Sorry! Image not available at this time

DarkMind: A new backdoor attack that leverages the reasoning capabilities of LLMs

techxplore.com - 17/Feb 15:41

Large language models (LLMs), such as the models supporting the functioning of ChatGPT, are now used by a growing number of people worldwide to source...