Taming Actor-Observer Asymmetry in Agents via Dialectical Alignment

2026-04-21

AI Summary

This study addresses actor-observer asymmetry (AOA)—a cognitive bias in multi-agent systems arising from role differences, which leads to inconsistent attributions and impaired fault diagnosis. It presents the first quantitative analysis and mitigation of AOA in large language model agents, introducing the ReTAS training framework. ReTAS integrates dialectical chain-of-thought reasoning with group-relative policy optimization, compelling agents to maintain perspective-invariant reasoning through self-reflection and mutual auditing, thereby achieving attribution consistency by synthesizing opposing viewpoints. Experimental results demonstrate that this approach significantly reduces attribution inconsistency in ambiguous failure scenarios, substantially improves fault resolution rates, and decreases the incidence of AOA effects by over 20%.

Technology Category

• Multiagent Systems: Adversarial Agents
• Cognitive Modeling & Cognitive Systems: Agent Architectures
• Knowledge Representation and Reasoning: Action, Change, and Causality

Application Category

• Search and Retrieval-Augmented AI: Agentic search
• Responsible Web: Machine-in-the-loop, human agency and autonomy
• User Modeling, Personalization and Recommendation: Fairness-aware retrieval and ranking

Abstract

Large Language Model agents have rapidly evolved from static text generators into dynamic systems capable of executing complex autonomous workflows. To enhance reliability, multi-agent frameworks assigning specialized roles are increasingly adopted to enable self-reflection and mutual auditing. While such role-playing effectively leverages domain expert knowledge, we find it simultaneously induces a human-like cognitive bias known as Actor-Observer Asymmetry (AOA). Specifically, an agent acting as an actor (during self-reflection) tends to attribute failures to external factors, whereas an observer (during mutual auditing) attributes the same errors to internal faults. We quantify this using our new Ambiguous Failure Benchmark, which reveals that simply swapping perspectives triggers the AOA effect in over 20% of cases for most models. To tame this bias, we introduce ReTAS (Reasoning via Thesis-Antithesis-Synthesis), a model trained through dialectical alignment to enforce perspective-invariant reasoning. By integrating dialectical chain-of-thought with Group Relative Policy Optimization, ReTAS guides agents to synthesize conflicting viewpoints into an objective consensus. Experiments demonstrate that ReTAS effectively mitigates attribution inconsistency and significantly improves fault resolution rates in ambiguous scenarios.

Problem

• Actor-Observer Asymmetry
• Large Language Model agents
• attribution bias
• multi-agent systems
• cognitive bias

Innovation

• Actor-Observer Asymmetry
• Dialectical Alignment
• ReTAS
• Ambiguous Failure Benchmark
• Multi-agent Reasoning