# Get Your Vitamin C! Robust Fact Verification with Contrastive Evidence

Source: [Get Your Vitamin C! Robust Fact Verification with Contrastive Evidence](<https://www.sota2.com/research/paper/68fa697b1e209131cea225a1>)

Collection index: [llms.txt](<https://www.sota2.com/research/llms.txt>)

Data updated: 2025-10-25T23:01:46.798Z

Authors: Tal Schuster, Adam Fisch, Regina Barzilay

Published: 2021-03-15T17:05:13.000Z

## Abstract

Typical fact verification models use retrieved written evidence to verify claims. Evidence sources, however, often change over time as more information is gathered and revised. In order to adapt, models must be sensitive to subtle differences in supporting evidence. We present VitaminC, a benchmark infused with challenging cases that require fact verification models to discern and adjust to slight factual changes. We collect over 100,000 Wikipedia revisions that modify an underlying fact, and leverage these revisions, together with additional synthetically constructed ones, to create a total of over 400,000 claim-evidence pairs. Unlike previous resources, the examples in VitaminC are contrastive, i.e., they contain evidence pairs that are nearly identical in language and content, with the exception that one supports a given claim while the other does not. We show that training using this design increases robustness -- improving accuracy by 10% on adversarial fact verification and 6% on adversarial natural language inference (NLI). Moreover, the structure of VitaminC leads us to define additional tasks for fact-checking resources: tagging relevant words in the evidence for verifying the claim, identifying factual revisions, and providing automatic edits via factually consistent text generation.

## Recorded benchmark results

A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage.

- [Claim Verification on DIALFACT (test)](<https://www.sota2.com/research/sota/claim-verification-on-dialfact-test.md>): Claim Verification; DIALFACT (test); Accuracy: 57.6.

- [Claim Verification on DIALFACT (val)](<https://www.sota2.com/research/sota/claim-verification-on-dialfact-val.md>): Claim Verification; DIALFACT (val); Accuracy: 60.5.

## Original sources and implementation

[Original paper](<https://arxiv.org/abs/2103.08541>)

[PDF](<https://arxiv.org/pdf/2103.08541.pdf>)