Pith. sign in

REVIEW 16 cited by

AgentInstruct: Toward Generative Teaching with Agentic Flows

Not yet reviewed by Pith; the record is open.

Re-run · record.json Download PDF Read on arXiv ↗

This paper has not been read by Pith yet. Machine review is queued; the pith claim, tier, and objections will appear here once it completes.

SPECIMEN: schema-true, not a live event

T0 review · schema-true

One-sentence machine reading of the paper's core claim.

pith:XXXXXXXX · record.json · timestamp

arxiv 2407.03502 v1 pith:7PDS7IYJ submitted 2024-07-03 cs.AI cs.CLcs.LG

Arindam Mitra , Luciano Del Corro , Guoqing Zheng , Shweti Mahajan , Dany Rouhana , Andres Codas , Yadong Lu , Wei-ge Chen
show 6 more authors
Olga Vrousgos Corby Rosset Fillipe Silva Hamed Khanpour Yash Lara Ahmed Awadallah
This is my paper
classification cs.AIcs.CLcs.LG
keywords dataimprovementmodelmodelssyntheticagentinstructcreatingagentic
verification ladder T0 review T1 audit T2 compute T3 formal
0 comments
read the original abstract

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.

Discussion (0). Sign in to comment.

Forward citations

Cited by 16 Pith papers

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. ISE: An Execution-Grounded Recipe for Multi-Turn OS-Agent Trajectories

    cs.CL 2026-06 conditional novelty 7.0 of 10

    ISE creates 23,132 execution-grounded multi-turn OS agent trajectories via intent simulation and live execution, improving agent performance on ClawEval from 19.3 to 37.7 pass@1 with Qwen3-8B.

  2. Masked Diffusion Language Models are Strong and Steerable Text-Based World Models for Agentic RL

    cs.AI 2026-05 conditional novelty 7.0 of 10

    Masked diffusion language models, not larger autoregressive LLMs, are the better building block for text-based world models in agentic RL, improving rollout fidelity, diversity, and downstream task success.

  3. Argus: A General-Purpose Agentic Runtime for Long-Horizon Reasoning

    cs.AI 2026-08 conditional novelty 6.0 of 10

    Argus demonstrates that a fixed-weight, self-evolving multi-role agentic runtime with verification-gated persistence can achieve competitive benchmark results and retain reusable state across long-horizon tasks.

  4. Search Hardness-Aware LLM-Based Problem Formulation for Expensive Simulation-Driven Design

    cs.NE 2026-07 conditional novelty 6.0 of 10

    SHA-PF uses initial simulation data to select a 'hard but promising' anchor satisfaction state and evolves LLM-generated formulations that prioritize it, reaching target designs with fewer expensive simulations on ant...

  5. Autodata: An agentic data scientist to create high quality synthetic data

    cs.AI 2026-06 unverdicted novelty 6.0 of 10

    Autodata trains meta-optimized AI agents to generate superior synthetic datasets, yielding performance gains over classical methods on CS research, legal, and math reasoning tasks.

  6. DataClaw0: Agentic Tailoring Multimodal Data from Raw Streams

    cs.LG 2026-06 unverdicted novelty 6.0 of 10

    DataClaw0 introduces an agentic data-tailoring paradigm, a 9B model trained on a synthetically generated dataset, and a new benchmark, claiming improved downstream adaptation in video generation, VQA, and GUI navigati...

  7. AgentSentinel: An End-to-End and Real-Time Security Defense Framework for Computer-Use Agents

    cs.CR 2025-09 conditional novelty 6.0 of 10

    AgentSentinel combines system-level tracing with LLM-based auditing to block 79.6% of attacks in the authors' 60-scenario computer-use agent benchmark.

  8. MUA-RL: Multi-turn User-interacting Agent Reinforcement Learning for agentic tool use

    cs.AI 2025-08 conditional novelty 6.0 of 10

    MUA-RL adds an LLM-simulated user to the RL rollout loop for multi-turn tool use, improving small Qwen3 models on TAU2, BFCL-V3 Multi Turn, and ACEBench Agent.

  9. CodeEvo: Interaction-Driven Synthesis of Code-centric Data through Hybrid and Iterative Feedback

    cs.SE 2025-07 conditional novelty 6.0 of 10

    CodeEvo uses two interacting LLM agents with keyword-guided instruction evolution and hybrid compiler-plus-LLM feedback to synthesize high-quality instruction-code pairs for fine-tuning code models.

  10. ScIRGen: Synthesize Realistic and Large-Scale RAG Dataset for Scientific Research

    cs.CL 2025-06 conditional novelty 6.0 of 10

    A framework and 61k QA dataset for geoscience RAG, generated from dataset metadata and papers with taxonomy-guided questions and filter-based answer validation.

  11. BioMol-MQA: A Multi-Modal Question Answering Dataset For LLM Reasoning Over Bio-Molecular Interactions

    cs.CL 2025-06 conditional novelty 6.0 of 10

    BioMol-MQA is a new multimodal QA dataset for polypharmacy in which LLMs perform poorly zero-shot but much better when given gold context.

  12. HunyuanOCR-1.5: Making Lightweight OCR VLMs Faster and Better

    cs.CV 2026-07 conditional novelty 5.0 of 10

    HunyuanOCR-1.5 makes a lightweight end-to-end OCR VLM the fastest among peers via DFlash speculative decoding while expanding long-tail capabilities through agent-driven data construction.

  13. Anchoring Refusal Direction: Mitigating Safety Risks in Tuning via Projection Constraint

    cs.CL 2025-09 conditional novelty 5.0 of 10

    ProCon anchors each sample's hidden-state projection onto the LLM's initial refusal direction during instruction fine-tuning, reducing refusal-direction drift and safety risks with limited task-performance loss.

  14. Hermes 4 Technical Report

    cs.AI 2025-08 conditional novelty 5.0 of 10

    Hermes 4 releases three open-weight reasoning models (14B, 70B, 405B) trained with synthetic data and a length-control SFT stage, evaluated on mathematics, code, knowledge, and alignment benchmarks.

  15. Breaking the Block: Preserving Data Continuity to Train Superior SAEs for Instruct Models

    cs.CL 2025-06 conditional novelty 5.0 of 10

    Training sparse autoencoders on individual chat sequences, instead of concatenated text blocks, improves reconstruction and feature interpretability on instruction-tuned LLMs.

  16. A Survey of LLM $\times$ DATA

    cs.DB 2025-05 conditional novelty 5.0 of 10

    A comprehensive survey of the bidirectional links between LLMs and data management, organized as DATA4LLM and LLM4DATA with a new 'IaaS' data-quality framework.

Pith tools