# The Behavior Gap: Evaluating Zero-shot LLM Agents in Complex Task-Oriented Dialogs

arXiv: https://arxiv.org/abs/2506.12266 · HTML: https://mlsanity.com/paper/2506.12266/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- Placed in General LLM agent frameworks and benchmarks, one of 16 directions in LLM agents and tool use.
- Builds on 37 of its 58 references that are in the map.
- First posted on arXiv 13 Jun 2025.

## Placed in

- LLM agents and tool use / [General LLM agent frameworks and benchmarks](https://mlsanity.com/agents/general-llm-agent-frameworks-and-benchmarks/)

## Builds on (most cited first, 30 of 37)

- [Language Models are Few-Shot Learners](https://mlsanity.com/paper/2005.14165/) (2020)
- [GPT-4 Technical Report](https://mlsanity.com/paper/2303.08774/) (2023)
- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [LLaMA: Open and Efficient Foundation Language Models](https://mlsanity.com/paper/2302.13971/) (2023)
- [Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks](https://mlsanity.com/paper/2005.11401/) (2020)
- [ReAct: Synergizing Reasoning and Acting in Language Models](https://mlsanity.com/paper/2210.03629/) (2022)
- [Judging LLM-as-a-Judge with MT-Bench and Chatbot Arena](https://mlsanity.com/paper/2306.05685/) (2023)
- [Qwen Technical Report](https://mlsanity.com/paper/2309.16609/) (2023)
- [G-Eval: NLG Evaluation using GPT-4 with Better Human Alignment](https://mlsanity.com/paper/2303.16634/) (2023)
- [A Survey on Large Language Model based Autonomous Agents](https://mlsanity.com/paper/2308.11432/) (2023)
- [The Rise and Potential of Large Language Model Based Agents: A Survey](https://mlsanity.com/paper/2309.07864/) (2023)
- [ChatDev: Communicative Agents for Software Development](https://mlsanity.com/paper/2307.07924/) (2023)
- [ExpeL: LLM Agents Are Experiential Learners](https://mlsanity.com/paper/2308.10144/) (2023)
- [DialoGPT: Large-Scale Generative Pre-training for Conversational Response Generation](https://mlsanity.com/paper/1911.00536/) (2019)
- [MultiWOZ -- A Large-Scale Multi-Domain Wizard-of-Oz Dataset for Task-Oriented Dialogue Modelling](https://mlsanity.com/paper/1810.00278/) (2018)
- [Towards a Human-like Open-Domain Chatbot](https://mlsanity.com/paper/2001.09977/) (2020)
- [Towards Scalable Multi-domain Conversational Agents: The Schema-Guided Dialogue Dataset](https://mlsanity.com/paper/1909.05855/) (2019)
- [MultiWOZ 2.2 : A Dialogue Dataset with Additional Annotation Corrections and State Tracking Baselines](https://mlsanity.com/paper/2007.12720/) (2020)
- [Unsupervised Evaluation of Interactive Dialog with DialoGPT](https://mlsanity.com/paper/2006.12719/) (2020)
- [Zhongjing: Enhancing the Chinese Medical Capabilities of Large Language Model through Expert Feedback and Real-world Multi-turn Dialogue](https://mlsanity.com/paper/2308.03549/) (2023)
- [Guiding Large Language Models via Directional Stimulus Prompting](https://mlsanity.com/paper/2302.11520/) (2023)
- [Dialog-based Interactive Image Retrieval](https://mlsanity.com/paper/1805.00145/) (2018)
- [Plug-and-Play Policy Planner for Large Language Model Powered Dialogue Agents](https://mlsanity.com/paper/2311.00262/) (2023)
- [SpokenWOZ: A Large-Scale Speech-Text Benchmark for Spoken Task-Oriented Dialogue Agents](https://mlsanity.com/paper/2305.13040/) (2023)
- [Prompt-Based Monte-Carlo Tree Search for Goal-Oriented Dialogue Policy Planning](https://mlsanity.com/paper/2305.13660/) (2023)
- [A User Simulator for Task-Completion Dialogues](https://mlsanity.com/paper/1612.05688/) (2016)
- [End-to-End Task-Completion Neural Dialogue Systems](https://mlsanity.com/paper/1703.01008/) (2017)
- [Shades of BLEU, Flavours of Success: The Case of MultiWOZ](https://mlsanity.com/paper/2106.05555/) (2021)
- [Grounding Gaps in Language Model Generations](https://mlsanity.com/paper/2311.09144/) (2023)

## Built on by (most cited first, 6 of 6)

- [Investigating LLM Variability in Personalized Conversational Information Retrieval](https://mlsanity.com/paper/2510.03795/) (2025)
- [Reasoning Gets Harder for LLMs Inside A Dialogue](https://mlsanity.com/paper/2603.20133/) (2026)
- [FloCA: Towards Faithful and Logically Consistent Flowchart Reasoning](https://mlsanity.com/paper/2602.14035/) (2026)
- [ESAinsTOD: A Unified End-to-End Schema-Aware Instruction-Tuning Framework for Task-Oriented Dialog Modeling](https://mlsanity.com/paper/2603.09691/) (2026)
- [When the Database Fails: Prompting LLM Dialogue Agents for Safe Recovery in Task-Oriented Dialogue](https://mlsanity.com/paper/2606.31307/) (2026)
- [Pretrained Persona Mixture Models and Tandem Models for Human Simulation](https://mlsanity.com/paper/2609.22607/) (2026)

## Abstract (arXiv)

Large Language Model (LLM)-based agents have significantly impacted Task-Oriented Dialog Systems (TODS) but continue to face notable performance challenges, especially in zero-shot scenarios. While prior work has noted this performance gap, the behavioral factors driving the performance gap remain under-explored. This study proposes a comprehensive evaluation framework to quantify the behavior gap between AI agents and human experts, focusing on discrepancies in dialog acts, tool usage, and knowledge utilization. Our findings reveal that this behavior gap is a critical factor negatively impacting the performance of LLM agents. Notably, as task complexity increases, the behavior gap widens (correlation: 0.963), leading to a degradation of agent performance on complex task-oriented dialogs. For the most complex task in our study, even the GPT-4o-based agent exhibits low alignment with human behavior, with low F1 scores for dialog acts (0.464), excessive and often misaligned tool usage with a F1 score of 0.139, and ineffective usage of external knowledge. Reducing such behavior gaps leads to significant performance improvement (24.3% on average). This study highlights the importance of comprehensive behavioral evaluations and improved alignment strategies to enhance the effectiveness of LLM-based TODS in handling complex tasks.