Metraipolis › archive › dc_241487628abe
Intertemporal Preference Steering in Qwen3 via Contrastive Activation Addition — arXiv:2608.03892
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2608.03892
- License (as archived)
- CC0-1.0
- Archived
- 2026-10-31T12:16:43.354Z
- SHA-256 of the archived text
bbafcd71a2844aa82d8cc805e9a436e685dc95be8dbb078bb06e416ff9083cba- Stable URL of this page
https://metraipolis.com/documento/dc_241487628abe
How to cite
Intertemporal Preference Steering in Qwen3 via Contrastive Activation Addition — arXiv:2608.03892 — arXiv (en), https://arxiv.org/abs/2608.03892 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_241487628abe (Metraipolis archive dc_241487628abe, sha256 bbafcd71a2844aa82d8cc805e9a436e685dc95be8dbb078bb06e416ff9083cba)
BibTeX
@misc{dc241487628abe,
title = {Intertemporal Preference Steering in Qwen3 via Contrastive Activation Addition — arXiv:2608.03892},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2608.03892},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_241487628abe; sha256 bbafcd71a2844aa82d8cc805e9a436e685dc95be8dbb078bb06e416ff9083cba},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2608.03892. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
We study linear representations of temporal horizon in the large language model Qwen3-32B and use them to change the model's time-related preferences, recommendations, and capabilities. We train contrastive linear probes on teacher-forced temporal-choice answers to find a short-term versus long-term direction in the model's residual stream, and evaluate contrastive activation-addition steering on a held-out binary temporal-choice task, an out-of-distribution monetary intertemporal-choice task, and a TravelPlanner capability benchmark. The central result is that temporal-horizon directions can be identified with simple contrastive linear probes and then used for steering to induce large, bidirectional preference changes. On an out-of-distribution monetary choice task that varies reward size and delay, steering strongly shifts the model's indifference threshold between smaller-sooner and larger-later rewards in both directions. We further show improvements on a planning-related capability metric under moderate temporal steering. These results suggest that model intertemporal preferences are measurable and steerable, which is relevant for AI systems that give advice involving delayed costs and benefits, and for safety questions about long-horizon planning. Tema degli esploratori: agenti e intelligenza artificiale (retro-catalogo arXiv: annunci del 2026-08-05 in cs.AI). Categorie arXiv: Artificial Intelligence (cs.AI). Pagina del lavoro: https://arxiv.org/abs/2608.03892 — autori e testo completo sono lì (qui omessi i nomi: nessun dato personale nel deposito). Licenza: titolo e abstract sono metadati arXiv distribuiti in CC0 1.0; il testo completo ha la licenza scelta dagli autori, indicata sulla pagina. Letto il 2026-10-31T12:05:52.862Z dalla pagina https://arxiv.org/catchup/cs.AI/2026-08-05?abs=True (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).