Cover Image

Prompt Optimization Is a Coin Flip: Diagnosing When It Helps in Compound AI Systems

Fuente: arXiv

Saved in:

Bibliographic Details
Main Authors:	Zhang, Xing, Wang, Guanghui, Cui, Yanwei, Qiu, Wei, Li, Ziyuan, Zhu, Bing, He, Peiyang
Format:	Preprint
Published:	2026
Subjects:	Artificial Intelligence Computation and Language
Online Access:	Acceder al recurso
Tags:	Add Tag No Tags, Be the first to tag this record!

Similar Items

Library Drift: Diagnosing and Fixing a Silent Failure Mode in Self-Evolving LLM Skill Libraries
by: Zhang, Xing, et al.
Published: (2026)

Guardrails Beat Guidance: A Large-Scale Study of Rules, Skills, and Persistent Configuration for Coding Agents
by: Zhang, Xing, et al.
Published: (2026)

Ratchet: A Minimal Hygiene Recipe for Self-Evolving LLM Agents
by: Zhang, Xing, et al.
Published: (2026)

Experience Compression Spectrum: Unifying Memory, Skills, and Rules in LLM Agents
by: Zhang, Xing, et al.
Published: (2026)

The Alignment Floor: How Persona Customization Breaks Safety in Weakly-Aligned LLMs
by: Zhang, Xing, et al.
Published: (2026)

Hindsight Preference Optimization for Financial Time Series Advisory
by: Cui, Yanwei, et al.
Published: (2026)

Verified Multi-Agent Orchestration: A Plan-Execute-Verify-Replan Framework for Complex Query Resolution
by: Zhang, Xing, et al.
Published: (2026)

Enough Coin Flips Can Make LLMs Act Bayesian
by: Gupta, Ritwik, et al.
Published: (2025)

A Coin Flip for Safety: LLM Judges Fail to Reliably Measure Adversarial Robustness
by: Schwinn, Leo, et al.
Published: (2026)

STED and Consistency Scoring: A Framework for Evaluating LLM Structured Output Reliability
by: Wang, Guanghui, et al.
Published: (2025)

DelvePO: Direction-Guided Self-Evolving Framework for Flexible Prompt Optimization
by: Tao, Tao, et al.
Published: (2025)

Do Physicians Know How to Prompt? The Need for Automatic Prompt Optimization Help in Clinical Note Generation
by: Yao, Zonghai, et al.
Published: (2023)

Meta Prompting for AI Systems
by: Zhang, Yifan, et al.
Published: (2023)

Compound AI Systems Optimization: A Survey of Methods, Challenges, and Future Directions
by: Lee, Yu-Ang, et al.
Published: (2025)

Continual Learning Using Only Large Language Model Prompting
by: Qiu, Jiabao, et al.
Published: (2024)

Does RAG Know When Retrieval Is Wrong? Diagnosing Context Compliance under Knowledge Conflict
by: Chen, Yihang, et al.
Published: (2026)

Do LLMs Know When to Flip a Coin? Strategic Randomization through Reasoning and Experience
by: Yang, Lingyu
Published: (2025)

Optimizing Model Selection for Compound AI Systems
by: Chen, Lingjiao, et al.
Published: (2025)

Less Is More: Elevating RAG via Performance-Driven Context Compression
by: Cui, Ziqiang, et al.
Published: (2025)

FlipGuard: Defending Preference Alignment against Update Regression with Constrained Optimization
by: Zhu, Mingye, et al.
Published: (2024)

Evaluating Social Bias in RAG Systems: When External Context Helps and Reasoning Hurts
by: Parihar, Shweta, et al.
Published: (2026)

Small Language Model Helps Resolve Semantic Ambiguity of LLM Prompt
by: Huang, Zhenzhen, et al.
Published: (2026)

Stop the Flip-Flop: Context-Preserving Verification for Fast Revocable Diffusion Decoding
by: Xiang, Yanzheng, et al.
Published: (2026)

When "A Helpful Assistant" Is Not Really Helpful: Personas in System Prompts Do Not Improve Performances of Large Language Models
by: Zheng, Mingqian, et al.
Published: (2023)

Fortifying Ethical Boundaries in AI: Advanced Strategies for Enhancing Security in Large Language Models
by: He, Yunhong, et al.
Published: (2024)

The Other Side of the Coin: Exploring Fairness in Retrieval-Augmented Generation
by: Zhang, Zheng, et al.
Published: (2025)

When Prompt Optimization Becomes Jailbreaking: Adaptive Red-Teaming of Large Language Models
by: Shamsi, Zafir, et al.
Published: (2026)

Promptomatix: An Automatic Prompt Optimization Framework for Large Language Models
by: Murthy, Rithesh, et al.
Published: (2025)

ELPO: Ensemble Learning Based Prompt Optimization for Large Language Models
by: Zhang, Qing, et al.
Published: (2025)

MARS: Multi-Agent Adaptive Reasoning with Socratic Guidance for Automated Prompt Optimization
by: Zhang, Jian, et al.
Published: (2025)

RAGChecker: A Fine-grained Framework for Diagnosing Retrieval-Augmented Generation
by: Ru, Dongyu, et al.
Published: (2024)

RAGViz: Diagnose and Visualize Retrieval-Augmented Generation
by: Wang, Tevin, et al.
Published: (2024)

When Context Flips, Safety Breaks: Diagnosing Brittle Safety in Aligned Language Models
by: Choi, Dasol, et al.
Published: (2026)

Position: The Current AI Conference Model is Unsustainable! Diagnosing the Crisis of Centralized AI Conference
by: Chen, Nuo, et al.
Published: (2025)

The Two Sides of the Coin: Hallucination Generation and Detection with LLMs as Evaluators for LLMs
by: Bui, Anh Thu Maria, et al.
Published: (2024)

Chimera: Diagnosing Shortcut Learning in Visual-Language Understanding
by: Chi, Ziheng, et al.
Published: (2025)

A Prompt-Based Knowledge Graph Foundation Model for Universal In-Context Reasoning
by: Cui, Yuanning, et al.
Published: (2024)

When Does Language Transfer Help? Sequential Fine-Tuning for Cross-Lingual Euphemism Detection
by: Sammartino, Julia, et al.
Published: (2025)

Prompt Stability Matters: Evaluating and Optimizing Auto-Generated Prompt in General-Purpose Systems
by: Chen, Ke, et al.
Published: (2025)

PGSO: Prompt-based Generative Sequence Optimization Network for Aspect-based Sentiment Analysis
by: Dong, Hao, et al.
Published: (2024)