mlsanityData to 14 Oct 2026

AgentInstruct: Toward Generative Teaching with Agentic Flows

arXiv 2407.03502 ↗First posted 3 Jul 2024cs.AI cs.CL cs.LGcited by 89

On the map: LLM agents and tool use → Educational agents and math problem solving

Abstract (arXiv)

Synthetic data is becoming increasingly important for accelerating the development of language models, both large and small. Despite several successful use cases, researchers also raised concerns around model collapse and drawbacks of imitating other models. This discrepancy can be attributed to the fact that synthetic data varies in quality and diversity. Effective use of synthetic data usually requires significant human effort in curating the data. We focus on using synthetic data for post-training, specifically creating data by powerful models to teach a new skill or behavior to another model, we refer to this setting as Generative Teaching. We introduce AgentInstruct, an extensible agentic framework for automatically creating large amounts of diverse and high-quality synthetic data. AgentInstruct can create both the prompts and responses, using only raw data sources like text documents and code files as seeds. We demonstrate the utility of AgentInstruct by creating a post training dataset of 25M pairs to teach language models different skills, such as text editing, creative writing, tool usage, coding, reading comprehension, etc. The dataset can be used for instruction tuning of any base model. We post-train Mistral-7b with the data. When comparing the resulting model Orca-3 to Mistral-7b-Instruct (which uses the same base model), we observe significant improvements across many benchmarks. For example, 40% improvement on AGIEval, 19% improvement on MMLU, 54% improvement on GSM8K, 38% improvement on BBH and 45% improvement on AlpacaEval. Additionally, it consistently outperforms other models such as LLAMA-8B-instruct and GPT-3.5-turbo.

Key facts. Cited by 89 papers in the map (data through 14 Oct 2026). Placed in Educational agents and math problem solving, one of 16 directions in LLM agents and tool use. Builds on 29 of its 62 references that are in the map. First posted on arXiv 3 Jul 2024.

Builds on

29 of 62 references.

References · 29

  1. 2021Training Verifiers to Solve Math Word Problemscited by 10,109
  2. 2021Measuring Mathematical Problem Solving With the MATH Datasetcited by 5,371
  3. 2018Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challengecited by 4,886
  4. 2023GPQA: A Graduate-Level Google-Proof Q&A Benchmarkcited by 2,879
  5. 2023AutoGen: Enabling Next-Gen LLM Applications via Multi-Agent Conversationcited by 1,988
  6. 2022Challenging BIG-Bench Tasks and Whether Chain-of-Thought Can Solve Themcited by 1,807
  7. 2023CAMEL: Communicative Agents for "Mind" Exploration of Large Language Model Societycited by 1,395
  8. 2023Instruction-Following Evaluation for Large Language Modelscited by 1,228
  9. 2023WizardLM: Empowering large pre-trained language models to follow complex instructionscited by 1,125
  10. 2019DROP: A Reading Comprehension Benchmark Requiring Discrete Reasoning Over Paragraphscited by 903
  11. 2023Enhancing Chat Language Models by Scaling High-quality Instructional Conversationscited by 839
  12. 2023MetaMath: Bootstrap Your Own Mathematical Questions for Large Language Modelscited by 735
17 more
  1. 2023AGIEval: A Human-Centric Benchmark for Evaluating Foundation Modelscited by 716
  2. 2023Instruction Tuning with GPT-4cited by 686
  3. 2023RLAIF vs. RLHF: Scaling Reinforcement Learning from Human Feedback with AI Feedbackcited by 625
  4. 2023Orca: Progressive Learning from Complex Explanation Traces of GPT-4cited by 379
  5. 2023The Curse of Recursion: Training on Generated Data Makes Models Forgetcited by 356
  6. 2024Benchmarking Retrieval-Augmented Generation for Medicinecited by 331
  7. 2023The False Promise of Imitating Proprietary LLMscited by 263
  8. 2023Camels in a Changing Climate: Enhancing LM Adaptation with Tulu 2cited by 229
  9. 2024Direct Nash Optimization: Teaching Language Models to Self-Improve with General Preferencescited by 178
  10. 2024Orca-Math: Unlocking the potential of SLMs in Grade School Mathcited by 151
  11. 2023ACI-BENCH: a Novel Ambient Clinical Intelligence Dataset for Benchmarking Automatic Visit Note Generationcited by 118
  12. 2023EQ-Bench: An Emotional Intelligence Benchmark for Large Language Modelscited by 80
  13. 2024FOFO: A Benchmark to Evaluate LLMs' Format-Following Capabilitycited by 80
  14. 2023Benchmarking Generation and Evaluation Capabilities of Large Language Models for Instruction Controllable Summarizationcited by 69
  15. 2020XtremeDistil: Multi-stage Distillation for Massive Multilingual Modelscited by 38
  16. 2024Autonomous Data Selection with Zero-shot Generative Classifiers for Mathematical Textscited by 27
  17. 2024Unearthing Large Scale Domain-Specific Knowledge from Public Corporacited by 12

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2024: 11; 2025: 45; 2026: 33.

89 papers.

Sort:

By area:

  1. 2025Kimi K2: Open Agentic Intelligencecited by 567
  2. 2025Multi-Agent Collaboration Mechanisms: A Survey of LLMscited by 431
  3. 2024ToolACE: Winning the Points of LLM Function Callingcited by 216
  4. 2025Phi-4-reasoning Technical Reportcited by 151
  5. 2025From LLM Reasoning to Autonomous AI Agents: A Comprehensive Reviewcited by 149
  6. 2025Reinforcement Learning for Long-Horizon Interactive LLM Agentscited by 108
  7. 2025Explorer: Scaling Exploration-driven Web Trajectory Synthesis for Multimodal Web Agentscited by 72
  8. 2025Distilling LLM Agent into Small Models with Retrieval and Code Toolscited by 47
  9. 2026On Data Engineering for Scaling LLM Terminal Capabilitiescited by 43
  10. 2024Unveiling the Secret Recipe: A Guide For Supervised Fine-Tuning Small LLMscited by 40
  11. 2026SkillGen: Verified Inference-Time Agent Skill Synthesiscited by 36
  12. 2025Fara-7B: An Efficient Agentic Model for Computer Usecited by 35
38 more
  1. 2026Tool-R0: Self-Evolving LLM Agents for Tool-Learning from Zero Datacited by 34
  2. 2025InSTA: Towards Internet-Scale Training For Agentscited by 32
  3. 2025MUA-RL: Multi-turn User-interacting Agent Reinforcement Learning for agentic tool usecited by 32
  4. 2025Self-Improving LLM Agents at Test-Timecited by 30
  5. 2026TermiGen: High-Fidelity Environment and Robust Trajectory Synthesis for Terminal Agentscited by 29
  6. 2025AgentSentinel: An End-to-End and Real-Time Security Defense Framework for Computer-Use Agentscited by 27
  7. 2026Gym-Anything: Turn any Software into an Agent Environmentcited by 27
  8. 2025DARS: Dynamic Action Re-Sampling to Enhance Coding Agent Performance by Adaptive Tree Traversalcited by 20
  9. 2025A Modular Approach for Clinical SLMs Driven by Synthetic Data with Pre-Instruction Tuning, Model Merging, and Clinical-Tasks Alignmentcited by 20
  10. 2024Automated test generation to evaluate tool-augmented LLMs as conversational AI agentscited by 17
  11. 2025Agent Data Protocol: Unifying Datasets for Diverse, Effective Fine-tuning of LLM Agentscited by 17
  12. 2025LoopTool: Closing the Data-Training Loop for Robust LLM Tool Callscited by 17
  13. 2026Autodata: An agentic data scientist to create high quality synthetic datacited by 16
  14. 2024CRAFT Your Dataset: Task-Specific Synthetic Dataset Generation Through Corpus Retrieval and Augmentationcited by 14
  15. 2025Krikri: Advancing Open Large Language Models for Greekcited by 14
  16. 2025SynthAgent: Adapting Web Agents with Synthetic Supervisioncited by 14
  17. 2026HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Bettercited by 14
  18. 2024TapeAgents: a Holistic Framework for Agent Development and Optimizationcited by 13
  19. 2024A Multi-AI Agent System for Autonomous Optimization of Agentic AI Solutions via Iterative Refinement and LLM-Driven Feedback Loopscited by 13
  20. 2026DIVE: Scaling Diversity in Agentic Task Synthesis for Generalizable Tool Usecited by 13
  21. 2024MAG-V: A Multi-Agent Framework for Synthetic Data Generation and Verificationcited by 12
  22. 2025Agentic AI Systems Applied to tasks in Financial Services: Modeling and model risk management crewscited by 12
  23. 2025ToolForge: A Data Synthesis Pipeline for Multi-Hop Search without Real-World APIscited by 10
  24. 2025A Strategic Coordination Framework of Small LLMs Matches Large LLMs in Data Synthesiscited by 9
  25. 2024Toward Secure Tuning: Mitigating Security Risks from Instruction Fine-Tuningcited by 8
  26. 2025Hermes 4 Technical Reportcited by 8
  27. 2026HumanLLM: Towards Personalized Understanding and Simulation of Human Naturecited by 8
  28. 2026TREX: Automating LLM Fine-tuning via Agent-Driven Tree-based Explorationcited by 8
  29. 2026Towards On-Policy Data Evolution for Visual-Native Multimodal Deep Search Agentscited by 8
  30. 2024YuLan-Mini: An Open Data-efficient Language Modelcited by 7
  31. 2025ScIRGen: Synthesize Realistic and Large-Scale RAG Dataset for Scientific Researchcited by 7
  32. 2025Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraintcited by 7
  33. 2025AgenticMath: Enhancing LLM Reasoning via Agentic-based Math Data Generationcited by 7
  34. 2025Adversarial D'ej`a Vu: Jailbreak Dictionary Learning for Stronger Generalization to Unseen Attackscited by 7
  35. 2026Collaborative Multi-Agent Scripts Generation for Enhancing Imperfect-Information Reasoning in Murder Mystery Gamescited by 6
  36. 2026Harness Engineering: Anatomy, Architecture, and Evolution of Coding Agents -- A Source-Code Study of Eleven Systemscited by 6
  37. 2025CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedbackcited by 5
  38. 2025GuidedSampling: Steering LLMs Towards Diverse Candidate Solutions at Inference-Timecited by 5

50 most cited of 89 All 89 (JSON)

Similar, not cited

Close in wording, no citation either way

  1. 2024Evaluating Language Models as Synthetic Data Generators
  2. 2026Mock Worlds, Real Skills: Building Small Agentic Language Models with Synthetic Tasks, Simulated Environments, and Rubric-Based Rewards
  3. 2025Synthetic Data Generation Using Large Language Models: Advances in Text and Code
  4. 2026SPT: Skills as Pre-Training Data for Agentic Language Models
  5. 2024Synthesizing Post-Training Data for LLMs through Multi-Agent Simulation
  6. 2026AutoTrainess: Teaching Language Models to Improve Language Models Autonomously
  7. 2022Unnatural Instructions: Tuning Language Models with (Almost) No Human Labor
  8. 2025BARE: Leveraging Base Language Models for Few-Shot Synthetic Data Generation