Towards Tracing Knowledge in Language Models Back to the Training Data

Ekin Akyurek, Tolga Bolukbasi, Frederick Liu, Binbin Xiong, Ian Tenney, Jacob Andreas, Kelvin Guu

Abstract
Language models (LMs) have been shown to memorize a great deal of factual knowledge contained in their training data. But when an LM generates an assertion, it is often difficult to determine where it learned this information and whether it is true. In this paper, we propose the problem of fact tracing: identifying which training examples taught an LM to generate a particular factual assertion. Prior work on training data attribution (TDA) may offer effective tools for identifying such examples, known as “proponents”. We present the first quantitative benchmark to evaluate this. We compare two popular families of TDA methods — gradient-based and embedding-based — and find that much headroom remains. For example, both methods have lower proponent-retrieval precision than an information retrieval baseline (BM25) that does not have access to the LM at all. We identify key challenges that may be necessary for further improvement such as overcoming the problem of gradient saturation, and also show how several nuanced implementation details of existing neural TDA methods can significantly improve overall fact tracing performance.
Anthology ID:
2022.findings-emnlp.180
Volume:
Findings of the Association for Computational Linguistics: EMNLP 2022
Month:
December
Year:
2022
Address:
Abu Dhabi, United Arab Emirates
Editors:
Yoav Goldberg, Zornitsa Kozareva, Yue Zhang
Venue:
Findings
SIG:
Publisher:
Association for Computational Linguistics
Note:
Pages:
2429–2446
Language:
URL:
https://aclanthology.org/2022.findings-emnlp.180
DOI:
10.18653/v1/2022.findings-emnlp.180
Bibkey:
Cite (ACL):
Ekin Akyurek, Tolga Bolukbasi, Frederick Liu, Binbin Xiong, Ian Tenney, Jacob Andreas, and Kelvin Guu. 2022. Towards Tracing Knowledge in Language Models Back to the Training Data. In Findings of the Association for Computational Linguistics: EMNLP 2022, pages 2429–2446, Abu Dhabi, United Arab Emirates. Association for Computational Linguistics.
Cite (Informal):
Towards Tracing Knowledge in Language Models Back to the Training Data (Akyurek et al., Findings 2022)
Copy Citation:
PDF:
https://preview.aclanthology.org/naacl-24-ws-corrections/2022.findings-emnlp.180.pdf
PDF Search