Built independently by an author, for readers. Read the story and support ChapterPal
ChapterPal Logo
Log inSign up

keyword

High-probability correctness

High-probability correctness is a statistical guarantee that an automated system or machine learning model produces outputs—such as predictions, factual assertions, or decision sets—that are accurate and valid with a predefined minimum level of confidence. Rather than relying on heuristic confidence scores or average-case performance metrics, this property establishes formal mathematical bounds ensuring that the probability of error does not exceed a chosen tolerance across the underlying data distribution. Often achieved through distribution-free uncertainty quantification techniques such as conformal prediction, high-probability correctness enables practitioners to calibrate risk thresholds, filter unreliable outputs, and enforce rigorous safety and factuality constraints in high-stakes algorithmic decision-making.

1 item

Language Models with Conformal Factuality Guarantees

Language Models with Conformal Factuality Guarantees

Christopher Mohri, Tatsunori Hashimoto

OrganizationsAllen Institute for AIUniversity of California, San Diego

Why you should read this

Proposes a conformal prediction framework that provides rigorous statistical correctness guarantees for black-box language model outputs by progressively pruning uncertain sub-claims while preserving most generated content.

Large language models demonstrate impressive capabilities across diverse tasks yet can generate content unfaithful to external knowledge sources.To enforce controllable generation,this work introduces post-generation techniques for constrained decoding crafted specifically as defensible guards.We introduce SENSE,a framework leveraging span-level evidence to align outputs with desired alignment criteria.Rather than enforcing token-level constraints directly over entire vocabularies,SENSE operates through differentiable integration within multimodal retrieval settings,and further demonstrates how regulatory signals interact.In extensive experiments spanning three datasets,two scientific benchmarks,guided narrative completion,and factor verification pipelines grounded via Wikidata concepts.

Added

2026-10-02