Co-Alignment: Rethinking Alignment as Bidirectional Human-AI Cognitive Adaptation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Li, Yubo, Song, Weiyi
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908660079263744
author Li, Yubo
Song, Weiyi
author_facet Li, Yubo
Song, Weiyi
contents Current AI alignment through RLHF follows a single directional paradigm that AI conforms to human preferences while treating human cognition as fixed. We propose a shift to co-alignment through Bidirectional Cognitive Alignment (BiCA), where humans and AI mutually adapt. BiCA uses learnable protocols, representation mapping, and KL-budget constraints for controlled co-evolution. In collaborative navigation, BiCA achieved 85.5% success versus 70.3% baseline, with 230% better mutual adaptation and 332% better protocol convergence. Emergent protocols outperformed handcrafted ones by 84%, while bidirectional adaptation unexpectedly improved safety (+23% out-of-distribution robustness). The 46% synergy improvement demonstrates optimal collaboration exists at the intersection, not union, of human and AI capabilities, validating the shift from single-directional to co-alignment paradigms.
format Preprint
id arxiv_https___arxiv_org_abs_2509_12179
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle Co-Alignment: Rethinking Alignment as Bidirectional Human-AI Cognitive Adaptation
Li, Yubo
Song, Weiyi
Artificial Intelligence
Multiagent Systems
Current AI alignment through RLHF follows a single directional paradigm that AI conforms to human preferences while treating human cognition as fixed. We propose a shift to co-alignment through Bidirectional Cognitive Alignment (BiCA), where humans and AI mutually adapt. BiCA uses learnable protocols, representation mapping, and KL-budget constraints for controlled co-evolution. In collaborative navigation, BiCA achieved 85.5% success versus 70.3% baseline, with 230% better mutual adaptation and 332% better protocol convergence. Emergent protocols outperformed handcrafted ones by 84%, while bidirectional adaptation unexpectedly improved safety (+23% out-of-distribution robustness). The 46% synergy improvement demonstrates optimal collaboration exists at the intersection, not union, of human and AI capabilities, validating the shift from single-directional to co-alignment paradigms.
title Co-Alignment: Rethinking Alignment as Bidirectional Human-AI Cognitive Adaptation
topic Artificial Intelligence
Multiagent Systems
url https://arxiv.org/abs/2509.12179