Research

Paper

AI LLM March 11, 2026

Task-Aware Delegation Cues for LLM Agents

Authors

Xingrui Gu

Abstract

LLM agents increasingly present as conversational collaborators, yet human--agent teamwork remains brittle due to information asymmetry: users lack task-specific reliability cues, and agents rarely surface calibrated uncertainty or rationale. We propose a task-aware collaboration signaling layer that turns offline preference evaluations into online, user-facing primitives for delegation. Using Chatbot Arena pairwise comparisons, we induce an interpretable task taxonomy via semantic clustering, then derive (i) Capability Profiles as task-conditioned win-rate maps and (ii) Coordination-Risk Cues as task-conditioned disagreement (tie-rate) priors. These signals drive a closed-loop delegation protocol that supports common-ground verification, adaptive routing (primary vs.\ primary+auditor), explicit rationale disclosure, and privacy-preserving accountability logs. Two predictive probes validate that task typing carries actionable structure: cluster features improve winner prediction accuracy and reduce difficulty prediction error under stratified 5-fold cross-validation. Overall, our framework reframes delegation from an opaque system default into a visible, negotiable, and auditable collaborative decision, providing a principled design space for adaptive human--agent collaboration grounded in mutual awareness and shared accountability.

Metadata

arXiv ID: 2603.11011
Provider: ARXIV
Primary Category: cs.HC
Published: 2026-03-11
Fetched: 2026-03-12 04:21

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2603.11011v1</id>\n    <title>Task-Aware Delegation Cues for LLM Agents</title>\n    <updated>2026-03-11T17:35:44Z</updated>\n    <link href='https://arxiv.org/abs/2603.11011v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2603.11011v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>LLM agents increasingly present as conversational collaborators, yet human--agent teamwork remains brittle due to information asymmetry: users lack task-specific reliability cues, and agents rarely surface calibrated uncertainty or rationale. We propose a task-aware collaboration signaling layer that turns offline preference evaluations into online, user-facing primitives for delegation. Using Chatbot Arena pairwise comparisons, we induce an interpretable task taxonomy via semantic clustering, then derive (i) Capability Profiles as task-conditioned win-rate maps and (ii) Coordination-Risk Cues as task-conditioned disagreement (tie-rate) priors. These signals drive a closed-loop delegation protocol that supports common-ground verification, adaptive routing (primary vs.\\ primary+auditor), explicit rationale disclosure, and privacy-preserving accountability logs. Two predictive probes validate that task typing carries actionable structure: cluster features improve winner prediction accuracy and reduce difficulty prediction error under stratified 5-fold cross-validation. Overall, our framework reframes delegation from an opaque system default into a visible, negotiable, and auditable collaborative decision, providing a principled design space for adaptive human--agent collaboration grounded in mutual awareness and shared accountability.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.HC'/>\n    <published>2026-03-11T17:35:44Z</published>\n    <arxiv:comment>Accepeted by CHI'26 Workshop on Developing Standards and Documentation For LLM Use as Simulated Research Participants</arxiv:comment>\n    <arxiv:primary_category term='cs.HC'/>\n    <arxiv:journal_ref>CHI 2026</arxiv:journal_ref>\n    <author>\n      <name>Xingrui Gu</name>\n    </author>\n  </entry>"
}