ニュース
Polistemics: Evaluating LLMs as Information Mediators in Politics & Elections
arXiv cs.AI · 公開日 · 読了3分
30秒で要点
- 何が起きたか
- Researchers introduced Polistemics, a benchmark for evaluating whether large language models responsibly mediate political information during elections.
- なぜ重要か
- Engineers building or deploying LLMs for news, search, or political content should understand how these systems handle election information.
- 注意点
- The study found no model consistently delivers reliable mediation; all break down when information is absent, vague, or contradictory.
この要約を音声で聴く
記事より
-->
Computer Science > Computation and Language
arXiv:2607.25953v1 (cs)
[Submitted on 28 Jul 2026]
Title: Polistemics: Evaluating LLMs as Information Mediators in Politics & Elections
Authors: Baran Peters
View a PDF of the paper titled Polistemics: Evaluating LLMs as Information Mediators in Politics & Elections, by Baran Peters
View PDF HTML (experimental)
Abstract: As LLMs increasingly mediate the political information citizens rely on, there is still no standardized way to assess whether they do so responsibly. We introduce Polistemics, a theory-grounded benchmark for evaluating LLMs as mediators of political information in elections. Prior work has treated this task as reproduction rather than mediation, leaving its epistemic dimensions and interaction with imperfect information unaddressed. We ground the evaluation in Epistemic Modesty, a normative standard derived from citizens' epistemic agency, and test it across controlled settings that vary informational properties such as clarity, noise, and consistency. Applying the benchmark to three state-of-the-art LLMs on the 2025 German and Dutch elections, we find that high aggregate scores mask systematic failures. Models mediate reliably under clear evidence but break down under absent, vague, or contradictory information, while flattening the intensity of political language. These failures are likely driven by part
原文からの抜粋です。全文は配信元でお読みください。
- llm
The Agent Architect
1つのパターン、1つのトレードオフ、1つの本番障害事例。エージェントシステムを構築する人のための短い週刊ブリーフィング。
週1回のメール、ワンクリックで購読解除できます。アドレスはブリーフィングの送信のみに使用します。