Foundational challenges in assuring alignment and safety of large language models
U Anwar, A Saparov, J Rando, D Paleka… - ar** the challenges of HCI: An application and evaluation of ChatGPT and GPT-4 for cost-efficient question answering
J Oppenlaender, J Hämäläinen - ar**_the_Challenges_of_HCI_An_Application_and_Evaluation_of_ChatGPT_and_GPT-4_for_Cost-Efficient_Question_Answering/links/675b39e9eea8d248be6470b9/Map**-the-Challenges-of-HCI-An-Application-and-Evaluation-of-ChatGPT-and-GPT-4-for-Cost-Efficient-Question-Answering.pdf" data-clk="hl=en&sa=T&oi=gga&ct=gga&cd=5&d=3945853756727773058&ei=zOarZ-eEM7zH6rQPz53buQk" data-clk-atid="gkdm8RV9wjYJ" target="_blank">[PDF] researchgate.net
[PDF][PDF] Map** the challenges of HCI: An application and evaluation of ChatGPT and GPT-4 for mining insights at scale
Large language models (LLMs), such as ChatGPT and GPT-4, are gaining wide-spread real
world use. Yet, these LLMs are closed source, and little is known about their performance in …
world use. Yet, these LLMs are closed source, and little is known about their performance in …
GBVSSL: Contrastive Semi-Supervised Learning Based on Generalized Bias-Variance Decomposition
S Li, L Han, Y Wang, J Zhu - Symmetry, 2024 - mdpi.com
Mainstream semi-supervised learning (SSL) techniques, such as pseudo-labeling and
contrastive learning, exhibit strong generalization abilities but lack theoretical …
contrastive learning, exhibit strong generalization abilities but lack theoretical …
Beyond Maximum Likelihood: Distribution-Aware Machine Learning
CJ Cundy - 2024 - search.proquest.com
Traditional machine learning approaches typically rely on maximum likelihood estimation
(MLE), due to its ease of implementation and equivalence to KL-divergence minimization …
(MLE), due to its ease of implementation and equivalence to KL-divergence minimization …
Position Paper: Dual-System Language Models via Next-Action Prediction
Z Du, WJ Su - ICML 2024 Workshop on LLMs and Cognition - openreview.net
In current Large Language Model (LLM) practices, each token is appended sequentially to
the output. In contrast, humans are capable of revising and correcting what we write …
the output. In contrast, humans are capable of revising and correcting what we write …
[PDF][PDF] Thoughtful, Confused, or Untrustworthy: How Text is Displayed Influences Perceptions of Generative AI Tools
Generative AI writing tools are rapidly being adopted in professional and creative writing,
offering writers the ability to generate and rewrite text in-and ex-situ. However, unexpected …
offering writers the ability to generate and rewrite text in-and ex-situ. However, unexpected …