paper-with-me

홈 › Papers

Using Synthetic Data to Train Neural Networks is Model-Based Reasoning

2017-03-02 · Tuan Anh Le, Atilim Gunes Baydin, Robert Zinkov, Frank Wood

We draw a formal connection between using synthetic training data to optimize neural network parameters and approximate, Bayesian, model-based reasoning. In particular, training a neural network using synthetic data can be viewed as learning a proposal distribution generator for approximate inference in the synthetic-data generative model. We demonstrate this connection in a recognition task where we develop a novel Captcha-breaking architecture and train it using synthetic data, demonstrating both state-of-the-art performance and a way of computing task-specific posterior uncertainty. Using a neural network trained this way, we also demonstrate successful breaking of real-world Captchas currently used by Facebook and Wikipedia. Reasoning from these empirical results and drawing connections with Bayesian modeling, we discuss the robustness of synthetic data results and suggest important considerations for ensuring good neural network generalization when training with synthetic data.

📄 PDF Abstract BibTeX arXiv:1703.00868

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

Enhancing Logical Reasoning in Large Language Models through Graph-based Synthetic Data

2024-09-19 · Jiaming Zhou, Abbas Ghaddar, Ge Zhang, Liheng Ma 외

Despite recent advances in training and prompting strategies for Large Language Models (LLMs), these models continue to face challenges with complex logical reasoning tasks that involve long reasoning chains. In this wor…

Logical ReasoningSpatial Reasoning

MIND: Math Informed syNthetic Dialogues for Pretraining LLMs

2024-10-15 · Syeda Nahida Akter, Shrimai Prabhumoye, John Kamalu, Sanjeev Satheesh 외

The utility of synthetic data to enhance pretraining data quality and hence to improve downstream task accuracy has been widely explored in recent large language models (LLMs). Yet, these approaches fall inadequate in co…

GSM8KMathMathematical ReasoningMMLU

CoT-Self-Instruct: Building high-quality synthetic prompts for reasoning and non-reasoning tasks

2025-07-31 · Ping Yu, Jack Lanchantin, Tianlu Wang, Weizhe Yuan 외 arxiv

We propose CoT-Self-Instruct, a synthetic data generation method that instructs LLMs to first reason and plan via Chain-of-Thought (CoT) based on given seed tasks, and then generate a new synthetic example of similar qua…

Synthetic Data Generation

ORACLE: Optimizing Reasoning Abilities of Large Language Models via Constraint-Led Synthetic Data Elicitation

2026-03-22 · Zhuojie Yang, Wentao Wan, Keze Wang arxiv

Training large language models (LLMs) with synthetic reasoning data has become a popular approach to enhancing their reasoning capabilities, while a key factor influencing the effectiveness of this paradigm is the qualit…

Learning from Synthetic Data Improves Multi-hop Reasoning

2026-03-02 · Anmol Kabra, Yilun Yin, Albert Gong, Kamilė Stankevičiūtė 외 arxiv

Reinforcement Learning (RL) has been shown to significantly boost reasoning capabilities of large language models (LLMs) in math, coding, and multi-hop reasoning tasks. However, RL fine-tuning requires abundant high-qual…

Reinforcement Learning