What Are You Doing? Effects of Intermediate Feedback from Agentic LLM In-Car Assistants During Multi-Step Processing
On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks
Abstract (arXiv)
Agentic AI assistants that autonomously perform multi-step tasks raise open questions for user experience: how should such systems communicate progress and reasoning during extended operations, especially in attention-critical contexts such as driving? We investigate feedback timing and verbosity from agentic LLM-based in-car assistants through a controlled, mixed-methods study (N=45) comparing planned steps and intermediate results feedback against silent operation with final-only response. Using a dual-task paradigm with an in-car voice assistant, we found that intermediate feedback significantly improved perceived speed, trust, and user experience while reducing task load - effects that held across varying task complexities and interaction contexts. Interviews further revealed user preferences for an adaptive approach: high initial transparency to establish trust, followed by progressively reducing verbosity as systems prove reliable, with adjustments based on task stakes and situational context. We translate our empirical findings into design implications for feedback timing and verbosity in agentic assistants, balancing transparency and efficiency.
Key facts. No citing papers in the map yet (data through 7 Oct 2026). Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use. Builds on 13 of its 87 references that are in the map. First posted on arXiv 17 Feb 2026.
Builds on
13 of 87 references.
Closest ancestors · 3
Same direction, newest first
- 2026CAR-bench: Evaluating the Consistency and Limit-Awareness of LLM Agents under Real-World Uncertainty29 Jan
- 2025Plan-Then-Execute: An Empirical Study of User Trust and Team Performance When Using LLM Agents As A Daily Assistant3 Feb
- 2023A Survey on Large Language Model based Autonomous Agents22 Aug
Other references · 10
- 2023Toolformer: Language Models Can Teach Themselves to Use Toolscited by 4,072
- 2023MemGPT: Towards LLMs as Operating Systemscited by 1,295
- 2024What Large Language Models Know and What People Think They Knowcited by 109
- 2023On the Calibration of Large Language Models and Alignmentcited by 72
- 2022Trust in Human-AI Interaction: Scoping Out Models, Measures, and Methodscited by 13
- 2025Mitigating Response Delays in Free-Form Conversations with LLM-powered Intelligent Virtual Agentscited by 12
- 2025CarMem: Enhancing Long-Term Memory in LLM Voice Assistants through Category-Boundingcited by 8
- 2019Robot Capability and Intention in Trust-based Decisions across Taskscited by 6
- 2023Integrating Generative Artificial Intelligence in Intelligent Vehicle Systemscited by 5
- 2026Agent2Agent Threats in Safety-Critical LLM Assistants: A Human-Centric Taxonomycited by 3
Built on by
Not cited yet.
Similar, not cited
Close in wording, no citation either way
- 2026TRACE: A Self-Evolving Skill Bank for Consistent, Limit-Aware LLM Agents
- 2026Feedback by Design: Understanding and Overcoming User Feedback Barriers in Conversational Agents
- 2025The Behavior Gap: Evaluating Zero-shot LLM Agents in Complex Task-Oriented Dialogs
- 2026Agentic Cognitive Depth: Operational Criteria for Evaluating LLM Agents
- 2026Less Interaction But More Explanation: A Communication Perspective on Agentic AI Interfaces
- 2026Probing an Embodied LLM: When Higher Observation Fidelity Hurts Problem Solving
- 2026When Should an AI Act? A Human-Centered Model of Scene, Context, and Behavior for Agentic AI Design
- 2025Beyond Task Completion: An Assessment Framework for Evaluating Agentic AI Systems