# A Framework for Evaluating LLMs Under Task Indeterminacy

arXiv: https://arxiv.org/abs/2411.13760 · HTML: https://mlsanity.com/paper/2411.13760/

## Key facts

- Cited by 3 papers in the map (data through 14 Oct 2026).
- Builds on 18 of its 48 references that are in the map.
- First posted on arXiv 21 Nov 2024.

## Builds on (most cited first, 18 of 18)

- [Measuring Massive Multitask Language Understanding](https://mlsanity.com/paper/2009.03300/) (2020)
- [GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understanding](https://mlsanity.com/paper/1804.07461/) (2018)
- [StereoSet: Measuring stereotypical bias in pretrained language models](https://mlsanity.com/paper/2004.09456/) (2020)
- [Red Teaming Language Models to Reduce Harms: Methods, Scaling Behaviors, and Lessons Learned](https://mlsanity.com/paper/2209.07858/) (2022)
- [Annotators with Attitudes: How Annotator Beliefs And Identities Bias Toxic Language Detection](https://mlsanity.com/paper/2111.07997/) (2021)
- [Who Validates the Validators? Aligning LLM-Assisted Evaluation of LLM Outputs with Human Preferences](https://mlsanity.com/paper/2404.12272/) (2024)
- [The 'Problem' of Human Label Variation: On Ground Truth in Data, Modeling and Evaluation](https://mlsanity.com/paper/2211.02570/) (2022)
- [On Releasing Annotator-Level Labels and Information in Datasets](https://mlsanity.com/paper/2110.05699/) (2021)
- [Red-Teaming for Generative AI: Silver Bullet or Security Theater?](https://mlsanity.com/paper/2401.15897/) (2024)
- [Stop Measuring Calibration When Humans Disagree](https://mlsanity.com/paper/2210.16133/) (2022)
- [A Safe Harbor for AI Evaluation and Red Teaming](https://mlsanity.com/paper/2403.04893/) (2024)
- [The Perspectivist Paradigm Shift: Assumptions and Challenges of Capturing Human Labels](https://mlsanity.com/paper/2405.05860/) (2024)
- [Like trainer, like bot? Inheritance of bias in algorithmic content moderation](https://mlsanity.com/paper/1707.01477/) (2017)
- [A Framework for Automated Measurement of Responsible AI Harms in Generative AI Applications](https://mlsanity.com/paper/2310.17750/) (2023)
- [D3CODE: Disentangling Disagreements in Data across Cultures on Offensiveness Detection and Evaluation](https://mlsanity.com/paper/2404.10857/) (2024)
- [Judgment Sieve: Reducing Uncertainty in Group Judgments through Interventions Targeting Ambiguity versus Disagreement](https://mlsanity.com/paper/2305.01615/) (2023)
- [Principled Evaluation with Human Labels: One Rater at a Time and Rater Equivalence](https://mlsanity.com/paper/2106.01254/) (2021)
- [Efficient Conformal Prediction via Cascaded Inference with Expanded Admission](https://mlsanity.com/paper/2007.03114/) (2020)

## Built on by (most cited first, 3 of 3)

- [Attention to Non-Adopters](https://mlsanity.com/paper/2510.15951/) (2025)
- [SCRuB: Social Concept Reasoning under Rubric-Based Evaluation](https://mlsanity.com/paper/2605.06444/) (2026)
- [PCS: Perceived Confidence Scoring of Black Box LLMs with Metamorphic Relations](https://mlsanity.com/paper/2502.07186/) (2025)

## Abstract (arXiv)

Large language model (LLM) evaluations often assume there is a single correct response -- a gold label -- for each item in the evaluation corpus. However, some tasks can be ambiguous -- i.e., they provide insufficient information to identify a unique interpretation -- or vague -- i.e., they do not clearly indicate where to draw the line when making a determination. Both ambiguity and vagueness can cause task indeterminacy -- the condition where some items in the evaluation corpus have more than one correct response. In this paper, we develop a framework for evaluating LLMs under task indeterminacy. Our framework disentangles the relationships between task specification, human ratings, and LLM responses in the LLM evaluation pipeline. Using our framework, we conduct a synthetic experiment showing that evaluations that use the "gold label" assumption underestimate the true performance. We also provide a method for estimating an error-adjusted performance interval given partial knowledge about indeterminate items in the evaluation corpus. We conclude by outlining implications of our work for the research community.