New Language Quantifies Uncertainty in LLM Applications.

Louis Mandel, Guillaume Baudart, Mandana Vaziri, Martin Hirzel· August 7, 2026 View original

Key takeaways

  • LLM applications struggle with reliability due to output uncertainty, especially in multi-step flows.
  • PPDL is a probabilistic language for programming LLM flows, quantifying and propagating uncertainty.
  • It allows experimentation with inference scaling without changing core application logic.
  • PPDL helps build more reliable and trustworthy LLM-based applications.

Who benefits

Software DevelopmentFinanceHealthcareLegalAI Research

Summary

PPDL introduces a probabilistic language for programming LLM-based flows, allowing developers to quantify and propagate uncertainty throughout an application. This enables experimentation with inference scaling techniques without altering the core logic, improving the reliability and trustworthiness of LLM applications.

Building reliable applications that integrate large language models (LLMs) presents a significant challenge, primarily due to the inherent uncertainty and lack of confidence measures in LLM outputs. This problem is compounded when multiple LLM calls and other tools are chained together in complex workflows, making it difficult for both developers and end-users to trust the final results. To address this, a new probabilistic language for programming LLM-based flows, called PPDL, has been introduced. PPDL empowers developers to explicitly quantify and propagate uncertainty throughout the entire application flow. This capability is crucial for understanding the reliability of the system's outputs at each stage. A key advantage of PPDL is its ability to facilitate experimentation with different inference scaling techniques without requiring any changes to the core logic of the flow. This separation allows developers to optimize performance and reliability more effectively. The paper demonstrates PPDL's capabilities through an experimental study and a case study involving the creation of a theorem-proving agent for the Rocq theorem prover, showcasing its practical utility in building more robust LLM-powered applications.

Why it matters

Professionals developing or deploying LLM-based applications need robust methods to manage and understand the uncertainty inherent in AI outputs. PPDL offers a way to build more reliable and trustworthy AI systems by explicitly modeling and propagating this uncertainty, which is critical for high-stakes applications.

How to implement this in your domain

  1. 1Explore PPDL or similar probabilistic programming paradigms for developing new LLM-based applications.
  2. 2Integrate uncertainty quantification into existing LLM workflows to improve transparency and reliability.
  3. 3Train development teams on probabilistic programming concepts and their application to AI systems.
  4. 4Experiment with different inference scaling techniques within LLM applications, leveraging tools that decouple logic from scaling.
  5. 5Develop internal standards for reporting confidence levels alongside LLM-generated outputs in critical applications.

Original post by Louis Mandel, Guillaume Baudart, Mandana Vaziri, Martin Hirzel

"arXiv:2608.05234v1 Announce Type: new Abstract: Building reliable applications that leverage large language models (LLMs) remains a significant challenge. While LLMs offer impressive capabilities across diverse tasks, their outputs often lack accuracy and provide no clear measure…"

View on X

Originally posted by Louis Mandel, Guillaume Baudart, Mandana Vaziri, Martin Hirzel on X · view source

Want to go deeper?

Turn these trends into skills with Learnijoy's hands-on AI & tech courses.

Explore courses