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...

Descrizione completa

Dettagli Bibliografici
Autori principali:	Kirk, HR, Vidgen, B, Röttger, P, Hale, SA
Natura:	Journal article
Lingua:	English
Pubblicazione:	Springer Nature 2024

Documenti analoghi

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

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

Cross-language Wikipedia editing of Okinawa, Japan
di: Hale, SA
Pubblicazione: (2015)

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

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

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

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

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

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

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

Morphological Alignment in Khorramabad Lori Language
di: Fatemeh Akoondi
Pubblicazione: (2023-03-01)

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

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

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