Weekly update
Weekly Newsletter
Excepteur sint occaecat cupidatat non proident
LLMs
Neurosymbolic AI could be leaner and smarter than today’s LLMs
Data-driven learning on a manifold (top). Data-driven learning and symbolic reasoning on a manifold (bottom). Conventional data-driven updates are followed with symbolic reasoning...
Lovabledaniels
Algorithm based on LLMs doubles lossless data compression rates
Image comparing the lossless compression rates of LMCompress with the traditional state-of-the-art methods and the large-model-based method that was proposed independently by a...
Lovabledaniels
Training LLMs to self-detoxify their language
Overview of SASA (self-disciplined autogressive sampling). Credit: arXiv (2024). DOI: 10.48550/arxiv.2410.03818 As we mature from childhood, our vocabulary—as well as the ways we...
Lovabledaniels
Small model approach could be more effective than LLMs
Credit: Unsplash/CC0 Public Domain Small language models are more reliable and secure than their large counterparts, primarily because they draw information from a...
Lovabledaniels
Researchers teach LLMs to solve complex planning challenges
Credit: Pixabay/CC0 Public Domain Imagine a coffee company trying to optimize its supply chain. The company sources beans from three suppliers, roasts them...
Lovabledaniels
‘Indiana Jones’ jailbreak approach highlights the vulnerabilities of existing LLMs
Example of how the jailbreaking approach works. Credit: Ding et al. Large language models (LLMs), such as the model underpinning the functioning of...
Lovabledaniels
A new backdoor attack that leverages the reasoning capabilities of LLMs
The user submits two queries (Q1 and Q2) to the backdoored customized LLM (the middle entity, highlighted in red). In Q1’s reasoning steps,...
Lovabledaniels
A fundamental revision of how AI acquires and processes language could result in more effective LLMs
Illustrative example of language acquisition through situation-based intention reading (left) and syntactico-semantic pattern finding (right), featuring a young child taking part in three...
Lovabledaniels
Tech startup proposes a novel way to tackle massive LLMs using the fastest memory available to mankind
GPU-like PCIe card offers 10PFLOPs FP4 compute power and 2GB of SRAM SRAM is usually used in small amounts as cache in processors...
Lovabledaniels
