Paraphrasing retrieved content is the most effective of five tested prompting defenses against domain-camouflaged injection attacks, cutting success rates 55-84% across three models while financial domains retain the highest residual risk.
hub Mixed citations
The Prompt Report: A Systematic Survey of Prompt Engineering Techniques
Mixed citation behavior. Most common role is background (60%).
44
Pith papers citing it
Background
60% of classified citations
abstract
Generative Artificial Intelligence (GenAI) systems are increasingly being deployed across diverse industries and research domains. Developers and end-users interact with these systems through the use of prompting and prompt engineering. Although prompt engineering is a widely adopted and extensively researched area, it suffers from conflicting terminology and a fragmented ontological understanding of what constitutes an effective prompt due to its relatively recent emergence. We establish a structured understanding of prompt engineering by assembling a taxonomy of prompting techniques and analyzing their applications. We present a detailed vocabulary of 33 vocabulary terms, a taxonomy of 58 LLM prompting techniques, and 40 techniques for other modalities. Additionally, we provide best practices and guidelines for prompt engineering, including advice for prompting state-of-the-art (SOTA) LLMs such as ChatGPT. We further present a meta-analysis of the entire literature on natural language prefix-prompting. As a culmination of these efforts, this paper presents the most comprehensive survey on prompt engineering to date.
hub tools
citation-role summary
background 4
method 1
citation-polarity summary
representative citing papers
Self-Harness lets LLM agents autonomously refine their interaction harnesses through weakness mining, proposal generation, and validation, raising held-out pass rates on Terminal-Bench-2.0 from 40.5% to 61.9%, 23.8% to 38.1%, and 42.9% to 57.1% across three models.
AtelierEval is the first unified benchmark that quantifies prompting proficiency of humans and MLLMs across 360 tasks using a cognitive taxonomy, with AtelierJudge providing scalable evaluation that correlates 0.79 with experts and shows mimicry outperforming planning.
TADI shows that domain-specialized tools orchestrated by an LLM over dual structured and semantic databases can convert heterogeneous wellsite data into evidence-grounded drilling intelligence, with tool design mattering more than model scale.
Vision-language models perform only marginally above random on action quality assessment and retain systematic biases even after targeted prompting and contrastive reformulation.
A systematic audit of LLM-based AI societies finds that 89.7% of 39 studies violate at least one of six PIMMUR validity principles, with reproductions showing that many claimed collective behaviors disappear when controls are tightened.
PromptCOS is a content-only watermarking method for LLM system prompts that embeds detectable cyclic signals via auxiliary tokens while preserving fidelity and resisting removal attacks.
PRIMETIME generator reveals that LLM datetime parsing and arithmetic primitives are individually unreliable but fully learnable via fine-tuning, enabling frontier-level accuracy on event planning with small LoRA models.
PromptGuard optimizes a universal safety soft prompt (and category-specific variants) in T2I embedding space to moderate NSFW inputs, achieving average unsafe ratios of 5.84-6.18% while being 3.8x faster than prior defenses.
Meta Agent Search uses a meta-agent to iteratively program novel agentic systems in code, producing agents that outperform state-of-the-art hand-designed ones across coding, science, and math while transferring across domains and models.
RECOM dataset shows automatic metrics for open-ended Reddit QA exhibit a validity-discrimination tradeoff, with cosine similarity strong on validity but weak on model ranking, and BERTScore showing the reverse pattern after length control.
Persona prompting trades expertise depth for reduced clarity in LLM answers and works best on advisory questions in medicine and psychology.
Transformer layers are analogous to power method steps, tilting tokens toward the principal eigenvector of the output-value weight product, with stronger analytical and empirical alignment in shared-weight models and a proposed steering method.
Intent Signal Theory formalizes four distinct intent-related objects in human-AI interaction, introduces a theorem on irreversible private intent loss, and reports supporting patterns from studies across LLMs, languages, and tasks.
Introduces TableGrid Navigation (TGN) and Progressive Inference Prompting (PIP) as training-free structured prompting frameworks that improve LLM performance on table question answering over baselines on TableBench and achieve SOTA on FeTaQa.
A dataset-agnostic framework converts text tool-calling benchmarks to paired audio evaluations via TTS, speaker variation and noise, then evaluates seven omni-modal models showing model- and task-dependent performance with small text-to-voice gaps.
Adapting multi-objective pure-exploration bandits enables efficient Pareto prompt set recovery and best feasible prompt identification for LLMs, with linear-case guarantees and empirical gains over baselines.
AI alignment must move beyond assuming users have fully formed goals and instead provide active cognitive support to help form and refine intent over time.
Arbiter-K is a governance-first architecture that turns probabilistic agent reasoning into discrete instructions with runtime taint propagation to block unsafe actions, reporting 76-95% interception rates and a 92.79% gain over baseline policies on two test systems.
LLMs improve with detailed code descriptions but remain insufficient to replace human annotators for security-specific qualitative coding.
Prompt Duel Optimizer uses dueling bandits and LLM-as-judge pairwise feedback with Double Thompson Sampling and top-performer mutation to find stronger prompts than label-free baselines on BBH and MS MARCO under limited comparison budgets.
FinKG-News constructs news-centric financial knowledge graphs to support in-context learning for credit risk report generation across three dimensions, claiming 19-34% quality gains and fewer hallucinations than baselines.
A taxonomy that consolidates prompt patterns from prior surveys into 30 unique canonical forms organized by two dimensions.
A 432-run experiment across capability tiers refutes the assumption of a monotone inverse relationship between LLM capability and optimal harness complexity, showing model-type-specific patterns instead.
citing papers explorer
-
Evaluating Prompting-Based Defenses Against Domain-Camouflaged Injection Attacks
Paraphrasing retrieved content is the most effective of five tested prompting defenses against domain-camouflaged injection attacks, cutting success rates 55-84% across three models while financial domains retain the highest residual risk.
-
Self-Harness: Harnesses That Improve Themselves
Self-Harness lets LLM agents autonomously refine their interaction harnesses through weakness mining, proposal generation, and validation, raising held-out pass rates on Terminal-Bench-2.0 from 40.5% to 61.9%, 23.8% to 38.1%, and 42.9% to 57.1% across three models.
-
AtelierEval: Agentic Evaluation of Humans & LLMs as Text-to-Image Prompters
AtelierEval is the first unified benchmark that quantifies prompting proficiency of humans and MLLMs across 360 tasks using a cognitive taxonomy, with AtelierJudge providing scalable evaluation that correlates 0.79 with experts and shows mimicry outperforming planning.
-
TADI: Tool-Augmented Drilling Intelligence via Agentic LLM Orchestration over Heterogeneous Wellsite Data
TADI shows that domain-specialized tools orchestrated by an LLM over dual structured and semantic databases can convert heterogeneous wellsite data into evidence-grounded drilling intelligence, with tool design mattering more than model scale.
-
Can Vision Language Models Judge Action Quality? An Empirical Evaluation
Vision-language models perform only marginally above random on action quality assessment and retain systematic biases even after targeted prompting and contrastive reformulation.
-
The PIMMUR Principles: Ensuring Validity in Collective Behavior of LLM Societies
A systematic audit of LLM-based AI societies finds that 89.7% of 39 studies violate at least one of six PIMMUR validity principles, with reproductions showing that many claimed collective behaviors disappear when controls are tightened.
-
PromptCOS: Towards Content-only System Prompt Copyright Auditing for LLMs
PromptCOS is a content-only watermarking method for LLM system prompts that embeds detectable cyclic signals via auxiliary tokens while preserving fidelity and resisting removal attacks.
-
PRIMETIME : Limits of LLMs in Temporal Primitives
PRIMETIME generator reveals that LLM datetime parsing and arithmetic primitives are individually unreliable but fully learnable via fine-tuning, enabling frontier-level accuracy on event planning with small LoRA models.
-
PromptGuard: Soft Prompt-Guided Unsafe Content Moderation for Text-to-Image Models
PromptGuard optimizes a universal safety soft prompt (and category-specific variants) in T2I embedding space to moderate NSFW inputs, achieving average unsafe ratios of 5.84-6.18% while being 3.8x faster than prior defenses.
-
Automated Design of Agentic Systems
Meta Agent Search uses a meta-agent to iteratively program novel agentic systems in code, producing agents that outperform state-of-the-art hand-designed ones across coding, science, and math while transferring across domains and models.
-
RECOM: A Validity Discrimination Tradeoff in Automatic Metrics for Open Ended Reddit Question Answering
RECOM dataset shows automatic metrics for open-ended Reddit QA exhibit a validity-discrimination tradeoff, with cosine similarity strong on validity but weak on model ranking, and BERTScore showing the reverse pattern after length control.
-
When Does Persona Prompting Actually Help? A Retrieval and Metric Analysis of Expert Role Injection in LLMs
Persona prompting trades expertise depth for reduced clarity in LLM answers and works best on advisory questions in medicine and psychology.
-
Analogies between Transformer Layers and Power Method
Transformer layers are analogous to power method steps, tilting tokens toward the principal eigenvector of the output-value weight product, with stronger analytical and empirical alignment in shared-weight models and a proposed steering method.
-
Intent Signal Theory: A Computational Framework for Intent-State Control in Human-AI Interaction
Intent Signal Theory formalizes four distinct intent-related objects in human-AI interaction, introduces a theorem on irreversible private intent loss, and reports supporting patterns from studies across LLMs, languages, and tasks.
-
Efficient Table QA via TableGrid Navigation and Progressive Inference Prompting
Introduces TableGrid Navigation (TGN) and Progressive Inference Prompting (PIP) as training-free structured prompting frameworks that improve LLM performance on table question answering over baselines on TableBench and achieve SOTA on FeTaQa.
-
From Text to Voice: A Reproducible and Verifiable Framework for Evaluating Tool Calling LLM Agents
A dataset-agnostic framework converts text tool-calling benchmarks to paired audio evaluations via TTS, speaker variation and noise, then evaluates seven omni-modal models showing model- and task-dependent performance with small text-to-voice gaps.
-
Efficient Multi-objective Prompt Optimization via Pure-exploration Bandits
Adapting multi-objective pure-exploration bandits enables efficient Pareto prompt set recovery and best feasible prompt identification for LLMs, with linear-case guarantees and empirical gains over baselines.
-
Alignment has a Fantasia Problem
AI alignment must move beyond assuming users have fully formed goals and instead provide active cognitive support to help form and refine intent over time.
-
From Craft to Kernel: A Governance-First Execution Architecture and Semantic ISA for Agentic Computers
Arbiter-K is a governance-first architecture that turns probabilistic agent reasoning into discrete instructions with runtime taint propagation to block unsafe actions, reporting 76-95% interception rates and a 92.79% gain over baseline policies on two test systems.
-
LLMs for Qualitative Data Analysis Fail on Security-specificComments in Human Experiments
LLMs improve with detailed code descriptions but remain insufficient to replace human annotators for security-specific qualitative coding.
-
LLM Prompt Duel Optimizer: Efficient Label-Free Prompt Optimization
Prompt Duel Optimizer uses dueling bandits and LLM-as-judge pairwise feedback with Double Thompson Sampling and top-performer mutation to find stronger prompts than label-free baselines on BBH and MS MARCO under limited comparison budgets.
-
Evidence-Supported Credit Risk Report Generation Using News-Centric Financial Knowledge Graphs
FinKG-News constructs news-centric financial knowledge graphs to support in-context learning for credit risk report generation across three dimensions, claiming 19-34% quality gains and fewer hallucinations than baselines.
-
A Taxonomy of Single-Turn Textual Prompt Patterns
A taxonomy that consolidates prompt patterns from prior surveys into 30 unique canonical forms organized by two dimensions.
-
It's Not the Capability: Harness Sensitivity Is Non-Monotone Across LLM Agent Tiers
A 432-run experiment across capability tiers refutes the assumption of a monotone inverse relationship between LLM capability and optimal harness complexity, showing model-type-specific patterns instead.
-
User Reviews as a Source for Usability Requirements: A Precursor Study on Using Large Language Models
LLMs can detect usability content in user reviews with F-scores comparable to humans, though performance depends strongly on prompt design.
-
LLARS: Enabling Domain Expert & Developer Collaboration for LLM Prompting, Generation and Evaluation
LLARS is a new integrated platform that combines collaborative prompt authoring, cost-controlled batch generation, and hybrid evaluation to help domain experts and developers jointly build and assess LLM systems.
-
U-Define: Designing User Workflows for Hard and Soft Constraints in LLM-Based Planning
U-Define improves user control in LLM planning by letting people define hard rules and soft preferences in natural language with matching verification methods, raising usefulness and satisfaction scores.
-
Looking Into the Past: Eye Movements Characterize Elements of Autobiographical Recall in Interviews with Holocaust Survivors
Eye movements during Holocaust survivor interviews vary by episodic, semantic, affective and temporal memory dimensions, with pre-onset gaze sufficient to predict sentence temporal context.
-
OOPrompt: Reifying Intents into Structured Artifacts for Modular and Iterative Prompting
OOPrompt reifies user intents into structured manipulable artifacts to enable modular and iterative prompting in LLM-based interactive systems.
-
Agent Mentor: Framing Agent Knowledge through Semantic Trajectory Analysis
Agent Mentor analyzes semantic trajectories in agent logs to identify undesired behaviors and derives corrective prompt instructions, yielding measurable accuracy gains on benchmark tasks across three agent setups.
-
Confidence Without Competence in AI-Assisted Knowledge Work
Standard LLM chats produce high perceived understanding but low objective learning in students, while future-self explanations best align confidence with actual gains and guided hints maximize learning with moderate workload.
-
The PICCO Framework for Large Language Model Prompting: A Taxonomy and Reference Architecture for Prompt Structure
PICCO is a five-element reference architecture (Persona, Instructions, Context, Constraints, Output) for structuring LLM prompts, derived from synthesizing prior frameworks along with a taxonomy distinguishing prompt concepts.
-
Self-Describing Structured Data with Dual-Layer Guidance: A Lightweight Alternative to RAG for Precision Retrieval in Large-Scale LLM Knowledge Navigation
SDSR places human metadata at file primacy and combines it with prompt routing rules to reach 100% primary category accuracy on a 119-category benchmark, far above the 65% no-guidance baseline.
-
Teaching Astronomy with Large Language Models
Structured integration of LLMs in astronomy education, including a domain-specific tutor and documentation requirements, leads to improved AI literacy and reduced student reliance on AI over the semester.
-
Comparing BERT Sentence-Pair Classification and Few-Shot LLM Prompting for Detecting Threat and Solution Framing in German Climate News
Fine-tuned BERT sentence-pair classifiers reach F1 0.83 while few-shot LLM prompting reaches F1 0.78 on threat and solution framing detection in 440 manually coded German climate news articles.
-
Characterizing Students' LLM Usage Behaviors and Their Association with Learning in Critical Thinking Tasks
Refined bottom-up categorization of LLM usage types in critical thinking homework, labeled by student initiative, shows associations with midterm performance across two course offerings.
-
Hint-Writing with Deferred AI Assistance: Fostering Critical Engagement in Data Science Education
In a randomized experiment with 97 graduate students, deferred AI assistance produced the highest-quality hints and helped students spot more code mistakes than independent writing or immediate AI help.
-
ClinQueryAgent: A Conversational Agent for Population Health Management
The paper introduces ClinQueryAgent, a conversational agent that converts natural language queries into database queries for population health management while keeping patient data secure, and reports its use by 128 staff across 15 NHS practices covering 148,319 patients.
-
Prompt Engineering Strategies for LLM-based Qualitative Coding of Psychological Safety in Software Engineering Communities: A Controlled Empirical Study
Multi-shot prompting raises agreement with humans for Claude Haiku but not DeepSeek-Chat or Gemini 2.5 Flash, with models showing different stability and a consistent bias toward over-labeling negative feedback.
-
CLaC at SemEval-2026 Task 6: Response Clarity Detection in Political Discourse
An LLM ensemble reached 80 macro-F1 on 3-class clarity detection and 59 on 9-class evasion detection, with partial layer unfreezing and multilingual ensembles improving encoder results while enriched context helped only LLMs.
-
A Reproducibility Study of Metacognitive Retrieval-Augmented Generation
MetaRAG is only partially reproducible with lower absolute scores than originally reported, gains substantially from reranking, and shows greater robustness than SIM-RAG under extended retrieval features.
-
The Prompt Engineering Report Distilled: Quick Start Guide for Life Sciences
The paper reduces a broad set of prompt engineering techniques to six core approaches and applies them to life sciences use cases while addressing common LLM pitfalls.
-
LLMs in Qualitative Research: Opportunities, Limitations, and Practical Considerations
The paper outlines opportunities, limitations, and practical parameters for integrating LLMs into qualitative research while aligning with epistemological commitments like reflexivity and interpretive judgment.
- MetaGraph: A Large-Scale Meta-Analysis of GenAI in Financial NLP (2022-2025)