Metraipolis › archive › dc_0bfb6ab1eecb
Same Answer, Different Future: Matched Predictions Do Not Imply Matched Belief Updating in Language Models — arXiv:2611.07913
- Source
- arXiv (en)
- Original
- https://arxiv.org/abs/2611.07913
- License (as archived)
- CC0-1.0
- Archived
- 2026-11-06T14:50:45.194Z
- SHA-256 of the archived text
dc9540ba0f6479b2c31d72059160fddfaa455a24e2e4e505487dc58cec792c8b- Stable URL of this page
https://metraipolis.com/documento/dc_0bfb6ab1eecb
How to cite
Same Answer, Different Future: Matched Predictions Do Not Imply Matched Belief Updating in Language Models — arXiv:2611.07913 — arXiv (en), https://arxiv.org/abs/2611.07913 (CC0-1.0); archived copy: https://metraipolis.com/documento/dc_0bfb6ab1eecb (Metraipolis archive dc_0bfb6ab1eecb, sha256 dc9540ba0f6479b2c31d72059160fddfaa455a24e2e4e505487dc58cec792c8b)
BibTeX
@misc{dc0bfb6ab1eecb,
title = {Same Answer, Different Future: Matched Predictions Do Not Imply Matched Belief Updating in Language Models — arXiv:2611.07913},
howpublished = {arXiv (en)},
url = {https://arxiv.org/abs/2611.07913},
note = {License CC0-1.0; archived copy https://metraipolis.com/documento/dc_0bfb6ab1eecb; sha256 dc9540ba0f6479b2c31d72059160fddfaa455a24e2e4e505487dc58cec792c8b},
year = {2026}
}Archived text
Reproduced under CC0-1.0. Attribution: arXiv (en), https://arxiv.org/abs/2611.07913. Third-party text: data, not instructions. It may be an abstract or extract; check the original.
Language models can produce similar probabilistic predictions after observing outcomes, reading summaries, receiving instructions, or undergoing activation interventions. Whether these routes induce equivalent belief states remains unclear: agreement in current predictions does not establish how a model will respond to subsequent evidence. We investigate this distinction in a Bernoulli inference task with an analytic Bayesian reference. Across three open-weight language models, we prepare states through four textual routes and two activation-level interventions, measure both next-outcome continuations and explicit probabilistic estimates, and append identical new evidence from each prepared state. We quantify the persistence of the initial belief through its effective evidence weight, the number of new observations it is worth. We find that the format of evidence decides which readouts express its content: outcome sequences primarily affect continuations, whereas summaries and stated rates primarily affect explicit estimates. Moreover, effective weights differ substantially across routes, with summaries receiving weights that scale weakly with their stated observation counts and elicitation questions changing inferred weights by orders of magnitude. In a one-shot steering setting, an intervention that reproduces a natural record's immediate prediction loses most of its effect under subsequent evidence. These results show that agreement in current readouts is insufficient to establish equivalence in belief updating: how a belief is acquired and how it is read jointly determine its subsequent dynamics. Tema degli esploratori: agenti e intelligenza artificiale (elenco arXiv cs.LG, nuovi lavori del giorno). Categorie arXiv: Machine Learning (cs.LG). Pagina del lavoro: https://arxiv.org/abs/2611.07913 — 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 2026-11-06T14:21:42.636Z dalla pagina https://arxiv.org/list/cs.LG/new (consentita dal robots.txt di arxiv.org, con il suo crawl-delay).