Imagen de Portada

Adapting the Interface, Not the Model: Runtime Harness Adaptation for Deterministic LLM Agents

Fuente: arXiv

Guardado en:

Detalles Bibliográficos
Autores principales:	Xu, Tianshi, Wen, Huifeng, Li, Meng
Formato:	Preprint
Publicado:	2026
Materias:	Artificial Intelligence
Acceso en línea:	Acceder al recurso
Etiquetas:	Agregar Etiqueta Sin Etiquetas, Sea el primero en etiquetar este registro!

Ejemplares similares

AI Harness Engineering: A Runtime Substrate for Foundation-Model Software Agents
por: Zhong, Hailin, et al.
Publicado: (2026)

SkillSmith: Compiling Agent Skills into Boundary-Guided Runtime Interfaces
por: Xu, Duling, et al.
Publicado: (2026)

DP-LLM: Runtime Model Adaptation with Dynamic Layer-wise Precision Assignment
por: Kwon, Sangwoo, et al.
Publicado: (2025)

Failure-Centered Runtime Evaluation for Deployed Trilingual Public-Space Agents
por: Meng, M.
Publicado: (2026)

Compiling Deterministic Structure into SLM Harnesses
por: Chong, Zan Kai, et al.
Publicado: (2026)

Harnessing LLM Agents with Skill Programs
por: Liu, Hongjun, et al.
Publicado: (2026)

HEQuant: Marrying Homomorphic Encryption and Quantization for Communication-Efficient Private Inference
por: Xu, Tianshi, et al.
Publicado: (2024)

Formal Skill: Programmable Runtime Skills for Efficient and Accurate LLM Agents
por: Zhang, Xi, et al.
Publicado: (2026)

AgentSpec: Customizable Runtime Enforcement for Safe and Reliable LLM Agents
por: Wang, Haoyu, et al.
Publicado: (2025)

Continual Harness: Online Adaptation for Self-Improving Foundation Agents
por: Karten, Seth, et al.
Publicado: (2026)

From Fragmentation to Integration: Exploring the Design Space of AI Agents for Human-as-the-Unit Privacy Management
por: Xu, Eryue, et al.
Publicado: (2026)

Stop Comparing LLM Agents Without Disclosing the Harness
por: Zhang, Yunbei, et al.
Publicado: (2026)

ProbGuard: Probabilistic Runtime Monitoring for LLM Agent Safety
por: Wang, Haoyu, et al.
Publicado: (2025)

Dark Patterns Meet GUI Agents: LLM Agent Susceptibility to Manipulative Interfaces and the Role of Human Oversight
por: Tang, Jingyu, et al.
Publicado: (2025)

Harness Updating Is Not Harness Benefit: Disentangling Evolution Capabilities in Self-Evolving LLM Agents
por: Lin, Minhua, et al.
Publicado: (2026)

FastQuery: Communication-efficient Embedding Table Query for Private LLM Inference
por: Lin, Chenqi, et al.
Publicado: (2024)

Robot Context Protocol (RCP): A Runtime-Agnostic Interface for Agent-Aware Robot Control
por: Lee, Lambert, et al.
Publicado: (2025)

Adapting LLM Agents with Universal Feedback in Communication
por: Wang, Kuan, et al.
Publicado: (2023)

RAP: Runtime Adaptive Pruning for LLM Inference
por: Liu, Huanrong, et al.
Publicado: (2025)

Operationalizing Data Minimization for Privacy-Preserving LLM Prompting
por: Zhou, Jijie, et al.
Publicado: (2025)

Harness-Bench: Measuring Harness Effects across Models in Realistic Agent Workflows
por: Yao, Yilun, et al.
Publicado: (2026)

Structured Personality Control and Adaptation for LLM Agents
por: Wang, Jinpeng, et al.
Publicado: (2026)

PrivCirNet: Efficient Private Inference via Block Circulant Transformation
por: Xu, Tianshi, et al.
Publicado: (2024)

Not My Agent, Not My Boundary? Elicitation of Personal Privacy Boundaries in AI-Delegated Information Sharing
por: Guo, Bingcan, et al.
Publicado: (2025)

Rescriber: Smaller-LLM-Powered User-Led Data Minimization for LLM-Based Chatbots
por: Zhou, Jijie, et al.
Publicado: (2024)

Autonomy Reshapes How Personalization Affects Privacy Concerns and Trust in LLM Agents
por: Zhang, Zhiping, et al.
Publicado: (2025)

A Methodology for Selecting and Composing Runtime Architecture Patterns for Production LLM Agents
por: Srinivasan, Vasundra
Publicado: (2026)

Runtime Governance for AI Agents: Policies on Paths
por: Kaptein, Maurits, et al.
Publicado: (2026)

General Modular Harness for LLM Agents in Multi-Turn Gaming Environments
por: Zhang, Yuxuan, et al.
Publicado: (2025)

AgentGuard: Runtime Verification of AI Agents
por: Koohestani, Roham
Publicado: (2025)

Secret Use of Large Language Model (LLM)
por: Zhang, Zhiping, et al.
Publicado: (2024)

Harnessing LLM to Attack LLM-Guarded Text-to-Image Models
por: Deng, Yimo, et al.
Publicado: (2023)

DLLM-Searcher: Adapting Diffusion Large Language Model for Search Agents
por: Zhao, Jiahao, et al.
Publicado: (2026)

Privacy Leakage Overshadowed by Views of AI: A Study on Human Oversight of Privacy in Language Model Agent
por: Zhang, Zhiping, et al.
Publicado: (2024)

Adapting Like Humans: A Metacognitive Agent with Test-time Reasoning
por: Li, Yang, et al.
Publicado: (2025)

Agentic LLMs as Powerful Deanonymizers: Re-identification of Participants in the Anthropic Interviewer Dataset
por: Li, Tianshi
Publicado: (2026)

Springdrift: An Auditable Persistent Runtime for LLM Agents with Case-Based Memory, Normative Safety, and Ambient Self-Perception
por: Brady, Seamus
Publicado: (2026)

ClawGuard: A Runtime Security Framework for Tool-Augmented LLM Agents Against Indirect Prompt Injection
por: Zhao, Wei, et al.
Publicado: (2026)

LLM-Hanabi: Evaluating Multi-Agent Gameplays with Theory-of-Mind and Rationale Inference in Imperfect Information Collaboration Game
por: Liang, Fangzhou, et al.
Publicado: (2025)

ANNEAL: Adapting LLM Agents via Governed Symbolic Patch Learning
por: Hakim, Safayat Bin, et al.
Publicado: (2026)