The benefits, risks and bounds of personalizing the alignment of large language models to individuals

The benefits, risks and bounds of personalizing the alignment of large language models to individuals

Large language models (LLMs) undergo ‘alignment’ so that they better reflect human values or preferences, and are safer or more useful. However, alignment is intrinsically difficult because the hundreds of millions of people who now interact with LLMs have different preferences for language and conv...

Fuld beskrivelse

Bibliografiske detaljer
Main Authors:	Kirk, HR, Vidgen, B, Röttger, P, Hale, SA
Format:	Journal article
Sprog:	English
Udgivet:	Springer Nature 2024

Lignende værker

Hatemoji: A test suite and adversarially-generated dataset for benchmarking and detecting emoji-based hate
af: Kirk, HR, et al.
Udgivet: (2022)

Is more data better? re-thinking the importance of efficiency in abusive language detection with transformers-based active learning
af: Kirk, HR, et al.
Udgivet: (2022)

Hatemoji: A test suite and adversarially-generated dataset for benchmarking and detecting emoji-based hate
af: Kirk, H, et al.
Udgivet: (2021)

Exploring large language models for ontology alignment
af: He, Y, et al.
Udgivet: (2023)

Survey on large language models alignment research
af: LIU Kunlin, et al.
Udgivet: (2024-06-01)

Survey on large language models alignment research
af: LIU Kunlin, et al.
Udgivet: (2024-06-01)

Two contrasting data annotation paradigms for subjective NLP tasks
af: Röttger, P, et al.
Udgivet: (2022)

Tackling racial bias in automated online hate detection: Towards fair and accurate detection of hateful users with geometric deep learning
af: Ahmed, Z, et al.
Udgivet: (2022)

Auditing large language models: a three-layered approach
af: Mökander, J, et al.
Udgivet: (2023)

Strong and weak alignment of large language models with human values
af: Mehdi Khamassi, et al.
Udgivet: (2024-08-01)

Aligning, autoencoding and prompting large language models for novel disease reporting
af: Liu, F, et al.
Udgivet: (2025)

HateCheck: functional tests for hate speech detection models
af: Röttger, P, et al.
Udgivet: (2021)

Deciphering implicit hate: evaluating automated detection algorithms for multimodal hate
af: Botelho, A, et al.
Udgivet: (2021)

Improvements in viral gene annotation using large language models and soft alignments
af: William L. Harrigan, et al.
Udgivet: (2024-04-01)

Personality prediction based on large language models
af: Wee, Jewel Xin Yu
Udgivet: (2024)

Individualizing the risks and benefits of postmenopausal hormone therapy.
af: van Staa, T, et al.
Udgivet: (2008)

Dyslexia application using axis align bounding boxes (AABB) /
af: Awatif Baharuddin, 1993-, et al.
Udgivet: (2016)

Evaluating the ability of large language models to emulate personality
af: Yilei Wang, et al.
Udgivet: (2025-01-01)

Directions in abusive language training data, a systematic review: Garbage in, garbage out.
af: Bertie Vidgen, et al.
Udgivet: (2020-01-01)

Large Language Model Augmentation and Feature Alignment Method for Few-Shot Continual Relation Extraction
af: LI Yifei, ZHANG Lingling, DONG Yuxuan, WANG Jiaxin, ZHONG Yujie, WEI Bifan
Udgivet: (2024-09-01)

Personalized prediction of lifetime benefits with statin therapy for asymptomatic individuals: a modeling study.
af: Bart S Ferket, et al.
Udgivet: (2012-01-01)

Person re-identification based on large vision-language model
af: Ding, Songyu
Udgivet: (2024)

Aligning Large Language Models for Enhancing Psychiatric Interviews Through Symptom Delineation and Summarization: Pilot Study
af: Jae-hee So, et al.
Udgivet: (2024-10-01)

DrugReAlign: a multisource prompt framework for drug repurposing based on large language models
af: Jinhang Wei, et al.
Udgivet: (2024-10-01)

Benefits of spontaneous confidence alignment between dyad members
af: Pescetelli, N, et al.
Udgivet: (2022)

Hedgerow benefits align with food production and sustainability goals
af: R Long, et al.
Udgivet: (2017-09-01)

Hedgerow benefits align with food production and sustainability goals
af: Rachael F. Long, et al.
Udgivet: (2017-09-01)

Alignment-free sequence comparison: benefits, applications, and tools
af: Andrzej Zielezinski, et al.
Udgivet: (2017-10-01)

Cross-language Wikipedia editing of Okinawa, Japan
af: Hale, SA
Udgivet: (2015)

Constructive alignment in a graduate-level project management course: an innovative framework using large language models
af: Estacio Pereira, et al.
Udgivet: (2024-04-01)

Affine Matching with Bounded Sensor Error: A Study of Geometric Hashing and Alignment
af: Grimson W. Eric L., et al.
Udgivet: (2004)

An upper bound on the convergence rate of a second functional in optimal sequence alignment
af: Hauser, R, et al.
Udgivet: (2017)

Dyslexia application using axis align bounding boxes (AABB) [electronic resource] /
af: Awatif Baharuddin, 1993-, et al.
Udgivet: (2016)

Outward Bound as education for personal growth
af: Katz, Richard, et al.
Udgivet: (2009)

The Existential Truth: On Personal beyond Bounds
af: K. S. Golikov
Udgivet: (2020-11-01)

The Opportunities and Risks of Large Language Models in Mental Health
af: Hannah R Lawrence, et al.
Udgivet: (2024-07-01)

Aligning subtitles in sign language videos
af: Bull, H, et al.
Udgivet: (2022)

Morphological Alignment in Khorramabad Lori Language
af: Fatemeh Akoondi
Udgivet: (2023-03-01)

Aligning English Language Testing With Curriculum
af: Marcela Palacio, et al.
Udgivet: (2016-07-01)

Impact of platform design on cross-language information exchange
af: Hale, SA
Udgivet: (2012)