Results 141 to 150 of about 176,171 (277)

LLMStinger: Jailbreaking LLMs using RL fine-tuned LLMs

We introduce LLMStinger, a novel approach that leverages Large Language Models (LLMs) to automatically generate adversarial suffixes for jailbreak attacks. Unlike traditional methods, which require complex prompt engineering or white-box access, LLMStinger uses a reinforcement learning (RL) loop to fine-tune an attacker LLM, generating new suffixes ...
Jha, Piyush, Arora, Arnav, Ganesh, Vijay
openaire   +2 more sources

Large‐Scale and Highly Reliable Hopfield Neural Networks Using Vertical NAND Flash Memory for the In‐Memory Associative Computing

Advanced Intelligent Systems, EarlyView.
Large‐scale Hopfield neural networks (HNNs) for associative computing are implemented using vertical NAND (VNAND) flash memory. The proposed VNAND HNN with the asynchronous update scenario achieve robust image restoration performance despite fabrication variations, while significantly reducing chip area (≈117× smaller than resistive random‐access ...
Jin Ho Chang, Jae Seung Woo, Suk‐Kang Sung, Ki‐Whan Song, Woo Young Choi   +4 more
wiley   +1 more source

When the Going Gets Weird, The Weird Turn Pro*: Management Best Practices in the Age of Medicinal Marijuana [PDF]

, 2020
Winn, John I., JD, LLM
core   +1 more source

A Generalizable Transformer Framework for Gene Regulatory Network Inference from Single‐Cell Transcriptomes

Advanced Intelligent Systems, EarlyView.
FTGRN introduces an LLM‐enhanced framework for gene regulatory network inference through a two‐stage workflow. It combines a Transformer‐based model, pretrained on GPT‐4 derived gene embeddings and regulatory knowledge, with a fine‐tuning stage utilizing single‐cell RNA‐seq data.
Guangzheng Weng, Hyobin Kim, Patrick Martin, Junil Kim, Tae‐Hyung Kim, Gi‐Hoon Nam, Dongha Kim, Kyoung Jae Won   +7 more
wiley   +1 more source

Low‐Temperature Reverse Water–Gas Shift Enabled by Magnetically Induced Catalysis

Angewandte Chemie, EarlyView.
Magnetically induced catalysis using standard Cu–Al spinel‐derived catalyst functionalized with carbon‐coated iron nanoparticles enables high CO yield (up to 62%) at mild catalyst and reactor temperatures (estimated at 300°C and determined as 25–123°C, respectively).
Junhui Hu, Lise Marie Lacroix, Jacob Johny, Sourav Ghosh, Elisabeth Hannah Wolf, Jeongmin Ji, Sheng‐Hsiang Lin, Manisha Durai, Alin Benice Schöne, Walid Hetaba, Holger Ruland, Walter Leitner, Alexis Bordet   +12 more
wiley   +2 more sources

LLM education

We propose a novel large language model (LLM) framework guided by GIS prior knowledge for generating professional GIS question banks.
openaire   +2 more sources

Predicting Materials Thermodynamics Enabled by Large Language Model‐Driven Dataset Building and Machine Learning

Advanced Intelligent Systems, EarlyView.
Illustration of text data mining of rare earth mineral thermodynamic parameters with the large language model‐powered LMExt. A dataset is built with mined thermodynamic properties. Subsequently, a machine learning model is trained to predict formation enthalpy from the dataset.
Juejing Liu, Haydn Anderson, Noah I. Waxman, Vsevolod Kovalev, Byron Fisher, Elizabeth Li, Xiaofeng Guo   +6 more
wiley   +1 more source

Evaluating LLMs at Detecting Errors in LLM Responses

With Large Language Models (LLMs) being widely used across various tasks, detecting errors in their responses is increasingly crucial. However, little research has been conducted on error detection of LLM responses. Collecting error annotations on LLM responses is challenging due to the subjective nature of many NLP tasks, and thus previous research ...
Kamoi, Ryo, Das, Sarkar Snigdha Sarathi, Lou, Renze, Ahn, Jihyun Janice, Zhao, Yilun, Lu, Xiaoxin, Zhang, Nan, Zhang, Yusen, Zhang, Ranran Haoran, Vummanthala, Sujeeth Reddy, Dave, Salika, Qin, Shaobo, Cohan, Arman, Yin, Wenpeng, Zhang, Rui   +14 more
openaire   +2 more sources

Robust Dysarthric Speech Recognition with GAN Enhancement and LLM Correction

Advanced Intelligent Systems, EarlyView.
This study tackles dysarthric speech recognition by combining generative adversarial network (GAN)‐generated synthetic data with large language model (LLM)‐based error correction. The approach integrates three key elements: an improved CycleGAN to generate synthetic dysarthric speech for data augmentation, a multimodal automatic speech recognition core
Yibo He, Kah Phooi Seng, Chee Shen Lim, Li Minn Ang   +3 more
wiley   +1 more source

LLM Augmented LLMs: Expanding Capabilities through Composition

Foundational models with billions of parameters which have been trained on large corpora of data have demonstrated non-trivial skills in a variety of domains. However, due to their monolithic structure, it is challenging and expensive to augment them or impart new skills.
Bansal, Rachit, Samanta, Bidisha, Dalmia, Siddharth, Gupta, Nitish, Vashishth, Shikhar, Ganapathy, Sriram, Bapna, Abhishek, Jain, Prateek, Talukdar, Partha   +8 more
openaire   +1 more source