# Truth-value judgment in language models: 'truth directions' are context sensitive

arXiv: https://arxiv.org/abs/2404.18865 · HTML: https://mlsanity.com/paper/2404.18865/

## Key facts

- Cited by 7 papers in the map (data through 14 Oct 2026).
- Builds on 12 of its 26 references that are in the map.
- First posted on arXiv 29 Apr 2024.

## Builds on (most cited first, 12 of 12)

- [Think you have Solved Question Answering? Try ARC, the AI2 Reasoning Challenge](https://mlsanity.com/paper/1803.05457/) (2018)
- [A large annotated corpus for learning natural language inference](https://mlsanity.com/paper/1508.05326/) (2015)
- [Inference-Time Intervention: Eliciting Truthful Answers from a Language Model](https://mlsanity.com/paper/2306.03341/) (2023)
- [Large Language Models Can Be Easily Distracted by Irrelevant Context](https://mlsanity.com/paper/2302.00093/) (2023)
- [Discovering Latent Knowledge in Language Models Without Supervision](https://mlsanity.com/paper/2212.03827/) (2022)
- [The Geometry of Truth: Emergent Linear Structure in Large Language Model Representations of True/False Datasets](https://mlsanity.com/paper/2310.06824/) (2023)
- [OLMo: Accelerating the Science of Language Models](https://mlsanity.com/paper/2402.00838/) (2024)
- [Towards Faithfully Interpretable NLP Systems: How should we define and evaluate faithfulness?](https://mlsanity.com/paper/2004.03685/) (2020)
- [Hypothesis Only Baselines in Natural Language Inference](https://mlsanity.com/paper/1805.01042/) (2018)
- [Explaining Answers with Entailment Trees](https://mlsanity.com/paper/2104.08661/) (2021)
- [Still No Lie Detector for Language Models: Probing Empirical and Conceptual Roadblocks](https://mlsanity.com/paper/2307.00175/) (2023)
- [Challenges with unsupervised LLM knowledge discovery](https://mlsanity.com/paper/2312.10029/) (2023)

## Built on by (most cited first, 7 of 7)

- [When Role-playing, Do Models Believe What They Say?](https://mlsanity.com/paper/2606.11502/) (2026)
- [Why Do LLMs Struggle in Strategic Play? Broken Links Between Observations, Beliefs, and Actions](https://mlsanity.com/paper/2605.00226/) (2026)
- [Testing the Limits of Truth Directions in LLMs](https://mlsanity.com/paper/2604.03754/) (2026)
- [Relational Linearity is a Predictor of Hallucinations](https://mlsanity.com/paper/2601.11429/) (2026)
- [LLM Probing with Contrastive Eigenproblems: Improving Understanding and Applicability of CCS](https://mlsanity.com/paper/2511.02089/) (2025)
- [Investigating Linear Probe Robustness to Linguistic Register, Medical Specialty, and Corpus Shifts in Medical QA](https://mlsanity.com/paper/2609.01361/) (2026)
- [Stress-Testing LLM Lie Detectors: Role-Play Failures and Spurious Correlations](https://mlsanity.com/paper/2609.39807/) (2026)

## Abstract (arXiv)

Recent work has demonstrated that the latent spaces of large language models (LLMs) contain directions predictive of the truth of sentences. Multiple methods recover such directions and build probes that are described as uncovering a model's "knowledge" or "beliefs". We investigate this phenomenon, looking closely at the impact of context on the probes. Our experiments establish where in the LLM the probe's predictions are (most) sensitive to the presence of related sentences, and how to best characterize this kind of sensitivity. We do so by measuring different types of consistency errors that occur after probing an LLM whose inputs consist of hypotheses preceded by (negated) supporting and contradicting sentences. We also perform a causal intervention experiment, investigating whether moving the representation of a premise along these truth-value directions influences the position of an entailed or contradicted sentence along that same direction. We find that the probes we test are generally context sensitive, but that contexts which should not affect the truth often still impact the probe outputs. Our experiments show that the type of errors depend on the layer, the model, and the kind of data. Finally, our results suggest that truth-value directions are causal mediators in the inference process that incorporates in-context information.