mlsanityData to 14 Oct 2026

AutoPDL: Automatic Prompt Optimization for LLM Agents

arXiv 2504.04365 ↗First posted 6 Apr 2025cs.LG cs.AI cs.PLcited by 27

On the map: LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

The performance of large language models (LLMs) depends on how they are prompted, with choices spanning both the high-level prompting pattern (e.g., Zero-Shot, CoT, ReAct, ReWOO) and the specific prompt content (instructions and few-shot demonstrations). Manually tuning this combination is tedious, error-prone, and specific to a given LLM and task. Therefore, this paper proposes AutoPDL, an automated approach to discovering good LLM agent configurations. Our approach frames this as a structured AutoML problem over a combinatorial space of agentic and non-agentic prompting patterns and demonstrations, using successive halving to efficiently navigate this space. We introduce a library implementing common prompting patterns using the PDL prompt programming language. AutoPDL solutions are human-readable, editable, and executable PDL programs that use this library. This approach also enables source-to-source optimization, allowing human-in-the-loop refinement and reuse. Evaluations across three tasks and seven LLMs (ranging from 3B to 70B parameters) show consistent accuracy gains (9.21±15.46 percentage points), up to 67.5pp, and reveal that selected prompting strategies vary across models and tasks.

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 33 of its 46 references that are in the map. First posted on arXiv 6 Apr 2025.

Builds on

33 of 46 references.

Closest ancestors · 1

Same direction, newest first

  1. 2024EvoAgent: Towards Automatic Multi-Agent Generation via Evolutionary Algorithms20 Jun

Other references · 32

  1. 2020Language Models are Few-Shot Learnerscited by 32,419
  2. 2024The Llama 3 Herd of Modelscited by 14,715
  3. 2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,459
  4. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  5. 2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,695
  6. 2022Large Language Models are Zero-Shot Reasonerscited by 4,374
  7. 2021Program Synthesis with Large Language Modelscited by 3,637
  8. 2022Beyond the Imitation Game: Quantifying and extrapolating the capabilities of language modelscited by 1,998
  9. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,988
  10. 2023Is Your Code Generated by ChatGPT Really Correct? Rigorous Evaluation of Large Language Models for Code Generationcited by 1,391
  11. 2018FEVER: a large-scale dataset for Fact Extraction and VERificationcited by 1,375
  12. 2022Large Language Models Are Human-Level Prompt Engineerscited by 1,045
20 more
  1. 2023DSPy: Compiling Declarative Language Model Calls into Self-Improving Pipelinescited by 832
  2. 2022PAL: Program-aided Language Modelscited by 652
  3. 2024Training Software Engineering Agents and Verifiers with SWE-Gymcited by 336
  4. 2023MINT: Evaluating LLMs in Multi-turn Interaction with Tools and Language Feedbackcited by 297
  5. 2024TextGrad: Automatic "Differentiation" via Textcited by 273
  6. 2015Non-stochastic Best Arm Identification and Hyperparameter Optimizationcited by 221
  7. 2023Efficient Guided Generation for Large Language Modelscited by 200
  8. 2023ReWOO: Decoupling Reasoning from Observations for Efficient Augmented Language Modelscited by 184
  9. 2023Black-Box Prompt Optimization: Aligning Large Language Models without Model Trainingcited by 122
  10. 2024Granite Code Models: A Family of Open Foundation Models for Code Intelligencecited by 117
  11. 2024Granite-Function Calling Model: Introducing Function Calling Abilities via Multi-task Learning of Granular Taskscited by 52
  12. 2025START: Self-taught Reasoner with Toolscited by 47
  13. 2024Prompt Optimization with EASE? Efficient Ordering-aware Automated Selection of Exemplarscited by 37
  14. 2024Prompt Optimization with Human Feedbackcited by 35
  15. 2024Aviary: training language agents on challenging scientific taskscited by 26
  16. 2021AlphaD3M: Machine Learning Pipeline Synthesiscited by 24
  17. 2024Localized Zeroth-Order Prompt Optimizationcited by 23
  18. 2016Automating biomedical data science through tree-based pipeline optimizationcited by 21
  19. 2024PDL: A Declarative Prompt Programming Languagecited by 14
  20. 2015Selecting Near-Optimal Learners via Incremental Data Allocationcited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 7; 2026: 20.

27 papers.

Sort:

By area:

  1. 2025Position: The Real Barrier to LLM Agent Usability is Agentic ROIcited by 12
  2. 2025Prompt-MII: Meta-Learning Instruction Induction for LLMscited by 9
  3. 2025SIPDO: Closed-Loop Prompt Optimization via Synthetic Data Feedbackcited by 7
  4. 2026MEMO: Memory-Augmented Model Context Optimization for Robust Multi-Turn Multi-Agent LLM Gamescited by 5
  5. 2025How can we assess human-agent interactions? Case studies in software agent designcited by 4
  6. 2026Learning to Rewrite Tool Descriptions for Reliable LLM-Agent Tool Usecited by 4
  7. 2026JTPRO: A Joint Tool-Prompt Reflective Optimization Framework for Language Agentscited by 4
  8. 2025Transduction is All You Need for Structured Data Workflowscited by 3
  9. 2026Experience as a Compass: Multi-agent RAG with Evolving Orchestration and Agent Promptscited by 3
  10. 2026Learning to Control LLM Agent Harnesses with Offline Reinforcement Learningcited by 3
  11. 2026Self-Optimizing Multi-Agent Systems for Deep Researchcited by 2
  12. 2026Lessons from the Field: An Adaptable Lifecycle Approach to Applied Dialogue Summarizationcited by 1
15 more
  1. 2026Resolving Java Code Repository Issues with iSWE Agentcited by 1
  2. 2026When Prompt Optimization Becomes Jailbreaking: Adaptive Red-Teaming of Large Language Modelscited by 1
  3. 2026A Reproducible Optimisation Protocol for Calibrating Prompt-Based Large Language Model Workflows in Evidence Synthesiscited by 1
  4. 2026MERA: Model Evolution and Routing with Skill Adaptation for Agentic Systems at Scalecited by 1
  5. 2025Representing Prompting Patterns with PDL: Compliance Agent Case Studycited by 0
  6. 2025Synthetic Data-Driven Prompt Tuning for Financial QA over Tables and Documentscited by 0
  7. 2026CROP: Token-Efficient Reasoning in Large Language Models via Regularized Prompt Optimizationcited by 0
  8. 2026Prompt Codebooks: Discrete Compositional Optimization for Language Model Instruction Refinementcited by 0
  9. 2026Automatic Extraction of Structured Information from Brain MRI Reports Using an Open-Weight Large Language Modelcited by 0
  10. 2026PPDL: LLM-Based Flows as Probabilistic Programscited by 0
  11. 2026StepReflect: Structured UI Transition Reflection for Mobile GUI Agentscited by 0
  12. 2026Code-to-Harness: Distilling Black-Box Optimizers from Self-Playcited by 0
  13. 2026Failure-Guided Co-Evolution of Prompts and Training Datacited by 0
  14. 2026From Weak Task Specifications to Scientific Extraction Agents: Optimizing Task Constructioncited by 0
  15. 2026Adaptive-GEPA: Make Your Harness Fit Heterogeneous Requestscited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2026Enhancing LLM Instruction Following: An Evaluation-Driven Multi-Agentic Workflow for Prompt Instructions Optimization
  2. 2025LLM-AutoDiff: Auto-Differentiate Any LLM Workflow
  3. 2026AutoLLMResearch: Training Research Agents for Automating LLM Experiment Configuration - Learning from Cheap, Optimizing Expensive
  4. 2023AgentTuning: Enabling Generalized Agent Abilities for LLMs
  5. 2024Designing Algorithms Empowered by Language Models: An Analytical Framework, Case Studies, and Insights
  6. 2024Minstrel: Structural Prompt Generation with Multi-Agents Coordination for Non-AI Experts
  7. 2024PRompt Optimization in Multi-Step Tasks (PROMST): Integrating Human Feedback and Heuristic-based Sampling
  8. 2025PRL: Prompts from Reinforcement Learning