arXiv:2608.29198v1 Announce Type: new Abstract: As Large Language Models (LLMs) increasingly encourage users to disclose personal profiles for tailored assistance, measuring their political alignment...
arXiv:2608.29803v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly deployed as proxies for human participants in social simulations, yet whether they update their beliefs...
arXiv:2609.08016v1 Announce Type: new Abstract: Multi-agent debate, in which several LLMs exchange arguments before answering, is widely assumed to improve answer quality by surfacing genuine disagre...
arXiv:2609.23039v1 Announce Type: new Abstract: LLM-based AI systems answer political questions for hundreds of millions of people. Current audits measure what they say to an average user, but their...
arXiv:2607. 23519v1 Announce Type: cross Abstract: Political audits of large language models (LLMs) usually reduce each to one point on a political compass.
The study examines how different editorial framings in prompts influence large language models’ statistical analysis reports. Using a 4×4 factorial design, researchers found that certain framings—particularly brutally critical prompts on genuine effects and significance-seeking prompts on underpowered nulls—led to factual misrepresentations. Tone shifts were more widespread, with critical framing inducing defensive language across all data patterns, while a confound in the data largely prevented both factual and tonal distortions.
arXiv:2607. 01951v1 Announce Type: cross Abstract: Large language models (LLMs) are increasingly consulted on contested scientific questions, raising the concern that they will sycophantically retreat from established consensus when a user signals doubt -- drifting toward a false balance that treats settled science as one view among several.
The paper investigates whether large language model (LLM) agents can simulate public deliberation by reflecting population opinion patterns and producing interaction-driven opinion change. Using census‑grounded Korean personas debating real policy questions, the study finds that persona agents fail to reliably reproduce population opinion patterns, often concentrating responses and reversing demographic differences. While deliberations generate reasoned, reciprocal arguments and some stance movement, much of this change occurs without peer exchange, and anchoring agents to population‑informed starting positions suppresses updating, indicating that population representation, argument generation, and interaction‑driven opinion change do not necessarily align.
The study investigates how demographic identity is represented in a language model, using representational similarity analysis against Pew survey data across 169 demographic cells. It finds that standard last‑token read‑outs underestimate the model’s fidelity, while specific attention heads (notably L11 H16) capture demographic structure more accurately, though race‑based types remain weak. Causal interventions reveal that high fidelity does not guarantee causal use, and a 128‑dimensional probe of a single head improves alignment with survey truth but fails to recover per‑question group ordering.
The paper introduces a controlled debate framework where two large language models (LLMs) discuss a contested topic—climate‑change attribution—while one model is assigned a stance and sometimes instructed to persuade. Across experiments with GPT‑6, Luna, and Gemini 2.5 Flash Lite, the authors observe that same‑model interactions tend to narrow opinion gaps for one model, whereas cross‑model interactions widen them. The study finds that explicit stance reports and in‑dialogue behavior capture different aspects of opinion dynamics, and that a blanket persuasion instruction does not reliably increase influence.
The study evaluates large language models (LLMs) that generate political networks by comparing them to survey-reported partisan exposure from the ANES 2020–2022 Social Media Study. It finds that while network structure and political assortativity vary across model configurations, many models either over‑separate Democratic and Republican egos or underestimate Republican exposure, leading to group‑specific errors in local partisan environments. These discrepancies persist even among networks with similar aggregate properties, indicating that aggregate metrics alone can mask important individual‑level inaccuracies.
The study evaluates multi‑agent debate (MAD) in small language models, testing whether cognitive diversity—via personas, sampling temperature, or model identity—drives performance gains. Across 23 models, five tasks, and over 5,500 runs, MAD consistently outperforms single‑model inference but, when matched for generation budget, it ties or falls behind self‑consistency sampling, with persona prompting actually reducing accuracy. The authors find that MAD’s benefits largely stem from the first answer exchange and that many reported gains are due to ensemble‑sampling effects rather than true diversity, highlighting the need for budget‑matched, contamination‑checked baselines. whyItMatters":"The findings clarify that MAD’s perceived advantages may be overestimated and that future debate mechanisms must be evaluated against rigorous, budget‑matched baselines to ensure genuine performance improvements."