The Trilemma of Truth in Large Language Models
Fuente:
arXiv
Saved in:
| Main Authors: | , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911266231025664 |
|---|---|
| author | Savcisens, Germans Eliassi-Rad, Tina |
| author_facet | Savcisens, Germans Eliassi-Rad, Tina |
| contents | The public often attributes human-like qualities to large language models (LLMs) and assumes they "know" certain things. In reality, LLMs encode information retained during training as internal probabilistic knowledge. This study examines existing methods for probing the veracity of that knowledge and identifies several flawed underlying assumptions. To address these flaws, we introduce sAwMIL (Sparse-Aware Multiple-Instance Learning), a multiclass probing framework that combines multiple-instance learning with conformal prediction. sAwMIL leverages internal activations of LLMs to classify statements as true, false, or neither. We evaluate sAwMIL across 16 open-source LLMs, including default and chat-based variants, on three new curated datasets. Our results show that (1) common probing methods fail to provide a reliable and transferable veracity direction and, in some settings, perform worse than zero-shot prompting; (2) truth and falsehood are not encoded symmetrically; and (3) LLMs encode a third type of signal that is distinct from both true and false. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2506_23921 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | The Trilemma of Truth in Large Language Models Savcisens, Germans Eliassi-Rad, Tina Computation and Language Machine Learning 68T50 I.2.6; I.2.7; G.3 The public often attributes human-like qualities to large language models (LLMs) and assumes they "know" certain things. In reality, LLMs encode information retained during training as internal probabilistic knowledge. This study examines existing methods for probing the veracity of that knowledge and identifies several flawed underlying assumptions. To address these flaws, we introduce sAwMIL (Sparse-Aware Multiple-Instance Learning), a multiclass probing framework that combines multiple-instance learning with conformal prediction. sAwMIL leverages internal activations of LLMs to classify statements as true, false, or neither. We evaluate sAwMIL across 16 open-source LLMs, including default and chat-based variants, on three new curated datasets. Our results show that (1) common probing methods fail to provide a reliable and transferable veracity direction and, in some settings, perform worse than zero-shot prompting; (2) truth and falsehood are not encoded symmetrically; and (3) LLMs encode a third type of signal that is distinct from both true and false. |
| title | The Trilemma of Truth in Large Language Models |
| topic | Computation and Language Machine Learning 68T50 I.2.6; I.2.7; G.3 |
| url | https://arxiv.org/abs/2506.23921 |