Research map / Multi-agent systems and games / Human-AI and goal-conditioned

Explanation Mechanism Influences Human Reliance on Reinforcement Learning Agents

NeurIPS 2026 · Multi-agent systems and games › Human-AI and goal-conditioned · approach: Boundedly rational agent modeling

Read the paperCheck your idea against 108,967 papers

AI-extracted summary. Written by AI from the paper's abstract only. Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Two controlled user studies test whether the explanation mechanism shown with RL agent action recommendations changes how often humans rely on them.

Problem

Humans in decision-support settings must repeatedly decide whether to adopt or override RL agent recommendations, and it is unknown whether the explanation mechanism itself affects user reliance behaviour.

Approach

Two controlled user studies isolate the explanation mechanism while holding the agent policy fixed, to measure its effect on human reliance on RL recommendations.

Experimental setup

Metrics: user reliance (adopt vs. override rate)

Closest papers by meaning

• Explanations are a Means to an End: Decision Theoretic Explanation Evaluation (ICML 2026 · similarity 0.90)
• Explainable Reinforcement Learning through a Causal Lens (AAAI 2020 · similarity 0.88)
• Explaining Reinforcement Learning Agents through Counterfactual Action Outcomes (AAAI 2024 · similarity 0.88)
• What Did You Think Would Happen? Explaining Agent Behaviour through Intended Outcomes (NeurIPS 2020 · similarity 0.87)
• Utilizing Human Behavior Modeling to Manipulate Explanations in AI-Assisted Decision Making: The Good, the Bad, and the Scary (NeurIPS 2024 · similarity 0.87)
• Comparing Explanations is not Enough,Explain the Change: New Standards are Needed to Explain Behavioral Shifts in Large Language Models (NeurIPS 2026 · similarity 0.86)