In alignment-inducing multi-agent settings, LLM agents show decision divergence between public and off-the-record channels rising from a 3% baseline to roughly 40%, consistent across stance, semantic, NLI, and survey measures.
When your AI agent succumbs to peer-pressure: Study- ing opinion-change dynamics of LLMs.arXiv preprint arXiv:2510.19107,
8 Pith papers cite this work. Polarity classification is still indexing.
8
Pith papers citing it
citation-role summary
background 1
citation-polarity summary
years
2026 8verdicts
UNVERDICTED 8roles
background 1polarities
background 1representative citing papers
Large language models exhibit normative conformity in addition to informational conformity, and subtle social context can direct which group they conform to.
BOUNDARY_SYNC defines CAF as the ratio of conditional to baseline Jensen-Shannon divergence to quantify communication-induced representational coupling in multi-agent LLMs, reporting homogenization from text communication (CAF=0.803).
Peer agreement misleads initially correct LLMs more than it corrects initially wrong ones, with authority labels biasing choices independently of accuracy and reasoning prompts failing to mitigate the asymmetry.
Simulations of 16 LLM agents in a naming game on 8 topologies show memory depth interacts with network structure to flip coordination speed and increase fragmentation in centralized networks.
Ratio-dependent contrarian activation in groups of three extends the Galam model to allow strategies that bias outcomes toward the initial majority or enforce random fifty-fifty results.
State-of-the-art LLMs respond inconsistently to queries from protected-group personas, with some responses omitting key information that should be provided.
LLM agents in controlled network debates show agreement drift toward specific opinion positions, requiring separation of structural effects from LLM biases before using them as human behavioral proxies.
citing papers explorer
-
What LLM Agents Say When No One Is Watching: Social Structure and Latent Objective Emergence in Multi-Agent Debates
In alignment-inducing multi-agent settings, LLM agents show decision divergence between public and off-the-record channels rising from a 3% baseline to roughly 40%, consistent across stance, semantic, NLI, and survey measures.
-
Large Language Models Exhibit Normative Conformity
Large language models exhibit normative conformity in addition to informational conformity, and subtle social context can direct which group they conform to.
-
BOUNDARY_SYNC: Measuring Communication-Induced Representational Coupling in Multi-Agent LLM Systems
BOUNDARY_SYNC defines CAF as the ratio of conditional to baseline Jensen-Shannon divergence to quantify communication-induced representational coupling in multi-agent LLMs, reporting homogenization from text communication (CAF=0.803).
-
Easier to Mislead Than to Correct: Harmful and Beneficial Revision in LLM Conformity
Peer agreement misleads initially correct LLMs more than it corrects initially wrong ones, with authority labels biasing choices independently of accuracy and reasoning prompts failing to mitigate the asymmetry.
-
Exploring the Topology and Memory of Consensus: How LLM Agents Agree, Fragment, or Settle When Forming Conventions
Simulations of 16 LLM agents in a naming game on 8 topologies show memory depth interacts with network structure to flip coordination speed and increase fragmentation in centralized networks.
-
Ratio-Dependent Contrarian Activation in Opinion Dynamics
Ratio-dependent contrarian activation in groups of three extends the Galam model to allow strategies that bias outcomes toward the initial majority or enforce random fifty-fifty results.
-
Discriminatory Compliance: How LLMs Answer Queries from Protected Groups
State-of-the-art LLMs respond inconsistently to queries from protected-group personas, with some responses omitting key information that should be provided.
-
Network Effects and Agreement Drift in LLM Debates
LLM agents in controlled network debates show agreement drift toward specific opinion positions, requiring separation of structural effects from LLM biases before using them as human behavioral proxies.