# HiL-Bench (Human-in-Loop Benchmark): Do Agents Know When to Ask for Help?

arXiv: https://arxiv.org/abs/2604.09408 · HTML: https://mlsanity.com/paper/2604.09408/

## Key facts

- Cited by 27 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 25 of its 37 references that are in the map.
- First posted on arXiv 10 Apr 2026.

## 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, 25 of 25)

- [Evaluating Large Language Models Trained on Code](https://mlsanity.com/paper/2107.03374/) (2021)
- [Program Synthesis with Large Language Models](https://mlsanity.com/paper/2108.07732/) (2021)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [WebArena: A Realistic Web Environment for Building Autonomous Agents](https://mlsanity.com/paper/2307.13854/) (2023)
- [LiveCodeBench: Holistic and Contamination Free Evaluation of Large Language Models for Code](https://mlsanity.com/paper/2403.07974/) (2024)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [ChatDev: Communicative Agents for Software Development](https://mlsanity.com/paper/2307.07924/) (2023)
- [Can LLM Already Serve as A Database Interface? A BIg Bench for Large-Scale Database Grounded Text-to-SQLs](https://mlsanity.com/paper/2305.03111/) (2023)
- [AmbigQA: Answering Ambiguous Open-domain Questions](https://mlsanity.com/paper/2004.10645/) (2020)
- [WorkArena: How Capable Are Web Agents at Solving Common Knowledge Work Tasks?](https://mlsanity.com/paper/2403.07718/) (2024)
- [TheAgentCompany: Benchmarking LLM Agents on Consequential Real World Tasks](https://mlsanity.com/paper/2412.14161/) (2024)
- [Asking Clarifying Questions in Open-Domain Information-Seeking Conversations](https://mlsanity.com/paper/1907.06554/) (2019)
- [Learning to Ask Good Questions: Ranking Clarification Questions using Neural Expected Value of Perfect Information](https://mlsanity.com/paper/1805.04655/) (2018)
- [Modeling Future Conversation Turns to Teach LLMs to Ask Clarifying Questions](https://mlsanity.com/paper/2410.13788/) (2024)
- [UserBench: An Interactive Gym Environment for User-Centric Agents](https://mlsanity.com/paper/2507.22034/) (2025)
- [QuestBench: Can LLMs ask the right question to acquire information in reasoning tasks?](https://mlsanity.com/paper/2503.22674/) (2025)
- [ConvAI3: Generating Clarifying Questions for Open-Domain Dialogue Systems (ClariQ)](https://mlsanity.com/paper/2009.11352/) (2020)
- [CLAMBER: A Benchmark of Identifying and Clarifying Ambiguous Information Needs in Large Language Models](https://mlsanity.com/paper/2405.12063/) (2024)
- [MCP-Atlas: A Large-Scale Benchmark for Tool-Use Competency with Real MCP Servers](https://mlsanity.com/paper/2602.00933/) (2026)
- [Interactive Code Generation via Test-Driven User-Intent Formalization](https://mlsanity.com/paper/2208.05950/) (2022)
- [Learning to Ask: When LLM Agents Meet Unclear Instruction](https://mlsanity.com/paper/2409.00557/) (2024)
- [Structured Uncertainty guided Clarification for LLM Agents](https://mlsanity.com/paper/2511.08798/) (2025)
- [Ask what's missing and what's useful: Improving Clarification Question Generation using Global Knowledge](https://mlsanity.com/paper/2104.06828/) (2021)
- [Teaching Language Models To Gather Information Proactively](https://mlsanity.com/paper/2507.21389/) (2025)

## Built on by (most cited first, 27 of 27)

- [AgentAbstain: Do LLM Agents Know When Not to Act?](https://mlsanity.com/paper/2607.10059/) (2026)
- [SWE-INTERACT: Reimagining SWE Benchmarks as User-Driven Long-Horizon Coding Sessions](https://mlsanity.com/paper/2606.30573/) (2026)
- [SWE-Touch: Benchmarking Coding Agents When Users Touch the Code](https://mlsanity.com/paper/2608.02499/) (2026)
- [What Makes Good Agentic Data? An ACE Lens on Data Generation for LLM Agents](https://mlsanity.com/paper/2608.27260/) (2026)
- [Beyond Oracle Communication: Benchmarking Interactive Intent Alignment Under Miscommunication and Evolving User Intent](https://mlsanity.com/paper/2609.38604/) (2026)
- [Don't Start What You Can't Finish: A Counterfactual Audit of Support-State Triage in LLM Agents](https://mlsanity.com/paper/2604.16752/) (2026)
- [Searching for Synergy in Shared Workspace Human-AI Collaboration](https://mlsanity.com/paper/2606.18413/) (2026)
- [REDDIT: Forgetting-Resistant Correction of Timestamp Drift in ASR via Replay-Based Distribution Editing](https://mlsanity.com/paper/2607.05364/) (2026)
- [LOGOS: A Living Logic for AI Agent Teams That Evolve With Humans](https://mlsanity.com/paper/2607.10878/) (2026)
- [Just A Rather Very Intelligent Spoken Agent](https://mlsanity.com/paper/2607.16610/) (2026)
- [Failure-Aware Long-Form Translation: Design and Implementation of a Recoverable LLM Translation System](https://mlsanity.com/paper/2608.09187/) (2026)
- [From Agent Behaviour to Agent-Friendly Documentation: An Empirical Study of How Coding Agents Discover, Read, and Write Technical Documentation](https://mlsanity.com/paper/2608.20195/) (2026)
- [Evaluating LLMs on Conversational Text-to-SQL under Chain Ambiguity and Intent Drift](https://mlsanity.com/paper/2608.29543/) (2026)
- [When Should a Failing Robot Ask? Initiating Corrective Human-Robot Dialogue from Audited Sensor Evidence](https://mlsanity.com/paper/2609.21942/) (2026)
- [XYEval: Agents say yes to bad advice](https://mlsanity.com/paper/2609.23939/) (2026)
- [Trust and Task Completion in the World of Consumer AI Agents](https://mlsanity.com/paper/2609.33017/) (2026)
- [What Happens During Autonomous Deep Research After the User Steps Away?](https://mlsanity.com/paper/2609.33509/) (2026)
- [Asking Earns Nothing: Scoring the Decision to Act in BFCL Multi-Turn](https://mlsanity.com/paper/2610.04429/) (2026)
- [From Evidence to Action: How Tool-Using Agents Fail](https://mlsanity.com/paper/2610.07753/) (2026)
- [Coding-Agent Benchmarks Should Match Their Users' Task Flows](https://mlsanity.com/paper/2610.09633/) (2026)
- [Conversational Task Disambiguation over Tabular Data: Leakage-Aware Formulation, Benchmark Suite, and Training](https://mlsanity.com/paper/2610.10740/) (2026)
- [TRACE: Diagnosing Verifier Brittleness in Agentic Evaluation](https://mlsanity.com/paper/2610.11678/) (2026)
- [CLEAT: Learning Value-Guided Clarification for Interactive Agents](https://mlsanity.com/paper/2610.12655/) (2026)
- [AutoEnvScaling: Automating the Data Flywheel with Terminal Agents](https://mlsanity.com/paper/2610.14163/) (2026)
- [FDE-Bench: Evaluating End-to-End Delivery from Underspecified Real-World Business Requests](https://mlsanity.com/paper/2610.14539/) (2026)
- [Instructions Are Evidence, Not Objectives: Measuring Outcome Alignment of Agents](https://mlsanity.com/paper/2610.17757/) (2026)
- [Agentic Self-Awareness: Aligning Knowing, Acting, and Holding](https://mlsanity.com/paper/2610.19226/) (2026)

## Abstract (arXiv)

Frontier coding agents solve complex tasks when given complete context but collapse when specifications are incomplete or ambiguous. The bottleneck is not raw capability, but judgment: knowing when to act autonomously and when to ask for help. Current benchmarks are blind to this failure mode. They supply unambiguous detailed instructions and solely reward execution correctness, so an agent that makes a lucky guess for a missing requirement will score identically to one that would have asked to be certain. We present HiL-Bench (Human-in-the-Loop Benchmark) to measure this selective escalation skill. Each task contains human-validated blockers (missing information, ambiguous requests, contradictory information) that surface only through progressive exploration, not upfront inspection. Our core metric, Ask-F1, the harmonic mean of question precision and blocker recall, captures the tension between over-asking and silent guessing; its structure architecturally prevents gaming through question spam. Evaluation across SWE and text-to-SQL domains reveals a large universal judgment gap: no frontier model recovers more than a fraction of its full-information performance when deciding whether to ask. Failure analysis identifies three key help-seeking patterns: overconfident wrong beliefs with no gap detection; high uncertainty detection yet persistent errors; broad, imprecise escalation without self-correction. These consistent patterns confirm poor help-seeking is a model-level flaw, not task-specific. RL training on shaped Ask-F1 reward shows judgment is trainable: a 32B model improves both help-seeking quality and task pass rate, with gains that transfer across domains. The model does not learn domain-specific heuristics for when to ask; it learns to detect unresolvable uncertainty and act on it.