REVIEW 5 cited by
Herd Behavior: Investigating Peer Influence in LLM-based Multi-Agent Systems
Not yet reviewed by Pith; the record is open.
This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.
SPECIMEN: schema-true, not a live event
T0 review · schema-true
One-sentence machine reading of the paper's core claim.
pith:XXXXXXXX · record.json · timestamp
read the original abstract
Recent advancements in Large Language Models (LLMs) have enabled the emergence of multi-agent systems where LLMs interact, collaborate, and make decisions in shared environments. While individual model behavior has been extensively studied, the dynamics of peer influence in such systems remain underexplored. In this paper, we investigate herd behavior, the tendency of agents to align their outputs with those of their peers, within LLM-based multi-agent interactions. We present a series of controlled experiments that reveal how herd behaviors are shaped by multiple factors. First, we show that the gap between self-confidence and perceived confidence in peers significantly impacts an agent's likelihood to conform. Second, we find that the format in which peer information is presented plays a critical role in modulating the strength of herd behavior. Finally, we demonstrate that the degree of herd behavior can be systematically controlled, and that appropriately calibrated herd tendencies can enhance collaborative outcomes. These findings offer new insights into the social dynamics of LLM-based systems and open pathways for designing more effective and adaptive multi-agent collaboration frameworks.
Forward citations
Cited by 5 Pith papers
-
The Evaluator Is Part of the Experiment: Measuring Open-Ended LLM Conformity
Open-ended LLM conformity is not captured by answer flips: wrong peers degrade revisions, and judges' ratings shift when peer context is visible, so evaluation must be modeled explicitly.
-
Most LLM Conformity Needs No Speaker: Measuring the Speaker-Free Floor in Peer-Pressure Benchmarks
Across six open-weight LLMs and seven datasets, a speaker-free wrong-answer assertion alone flips 66.5% of initially correct answers, versus 10.3% for a plain re-ask; source labels mainly add a modest increment above ...
-
Probing Multimodal Large Language Models on Cognitive Biases in Chinese Short-Video Misinformation
Multimodal LLMs exhibit different levels of susceptibility to misinformation in short videos, with Gemini-2.5-Pro showing the highest resistance (belief score 71.5) and o3 the lowest (35.2).
-
Social Networks of LLM Agents
Attention width and source social power determine whether LLM agent networks herd or achieve wisdom-of-crowds, with a pricing equalizer restoring optimal collective weights.
-
MAD-Spear: A Conformity-Driven Prompt Injection Attack on Multi-Agent Debate Systems
MAD-Spear is a prompt injection attack that makes a single compromised agent emit fake 'Sybil' peer answers, exploiting LLM conformity to steer a multi-agent debate toward a wrong consensus and higher token costs.
Discussion (0). Sign in to comment.