Towards Tracing Knowledge in Language Models Back to the Training Data (2022.findings-emnlp)
Copied to clipboard
| Challenge: | Prior work on training data attribution (TDA) may offer effective tools for identifying such examples, known as "proponents". |
| Approach: | They propose a benchmark to identify which training examples taught an LM to generate a particular factual assertion. |
| Outcome: | The proposed methods have lower proponent-retrieval precision than baselines that do not have access to the LM. |
Similar Papers
Revealing the Parametric Knowledge of Language Models: A Unified Framework for Attribution Methods (2024.acl-long)
Copied to clipboard
| Challenge: | Language Models acquire parametric knowledge from their training process, embedding it within their weights. |
| Approach: | They propose a new evaluation framework to quantify and compare the knowledge revealed by Instance Attribution and Neuron Attributions. |
| Outcome: | The proposed evaluation framework compares the knowledge revealed by IA and NA with that of neuron attribution methods. |
Give Me the Facts! A Survey on Factual Knowledge Probing in Pre-trained Language Models (2023.findings-emnlp)
Copied to clipboard
| Challenge: | Pre-trained language models are trained on vast unlabeled data, rich in world knowledge. |
| Approach: | They propose a categorization scheme for factual probing methods based on how inputs, outputs and probed PLMs are adapted . they synthesize insights about knowledge retention and prompt optimization in PLM models and analyze obstacles to adopting them as knowledge bases . |
| Outcome: | The proposed method synthesizes insights about knowledge retention and prompt optimization in PLMs, analyzes obstacles to adopting them as knowledge bases and outline directions for future work. |
Information-Guided Identification of Training Data Imprint in (Proprietary) Large Language Models (2025.naacl-long)
Copied to clipboard
Abhilasha Ravichander, Jillian Fisher, Taylor Sorensen, Ximing Lu, Maria Antoniak, Bill Yuchen Lin, Niloofar Mireshghallah, Chandra Bhagavatula, Yejin Choi
| Challenge: | Lack of transparency in training data is limiting external oversight and inspection of LLMs for issues such as copyright infringement and data contamination. |
| Approach: | They propose a method to identify training data known to proprietary LLMs without requiring access to model weights or token probabilities by using information-guided probes. |
| Outcome: | The proposed method can identify training data known to proprietary LLMs without access to model weights or token probabilities. |
Can LMs Learn New Entities from Descriptions? Challenges in Propagating Injected Knowledge (2023.acl-long)
Copied to clipboard
| Challenge: | Existing methods for updating knowledge show little propagation of injected knowledge. |
| Approach: | They propose to inject individual facts into LMs and evaluate whether they can propagate injected facts while not changing predictions on other contexts. |
| Outcome: | The proposed model can make inferences based on injected facts and propagate them . existing methods show little propagation of injected knowledge . |
How Can We Know What Language Models Know? (2020.tacl-1)
Copied to clipboard
| Challenge: | Recent work examines knowledge contained in language models by having the LM fill in the blanks of prompts such as “Obama is a __ by profession”. |
| Approach: | They propose mining-based and paraphrasing-based methods to automatically generate high-quality and diverse prompts, as well as ensemble methods to combine answers from different prompts. |
| Outcome: | The proposed methods improve accuracy from 31.1% to 39.6% on the LAMA benchmark for extracting relational knowledge from LMs. |
Robust Data Watermarking in Language Models by Injecting Fictitious Knowledge (2025.findings-acl)
Copied to clipboard
| Challenge: | Data watermarking in language models injects traceable signals, such as specific token sequences or stylistic patterns, into copyrighted text, allowing copyright holders to track and verify training data ownership. |
| Approach: | They propose a data watermarking approach that injects coherent and plausible yet fictitious knowledge into training data using generated passages describing a fictious entity and its associated attributes. |
| Outcome: | The proposed method is designed to be memorized by the LLM, and that increasing their density, length, and diversity of attributes strengthens their memorization. |
Unlearning Traces the Influential Training Data of Language Models (2024.acl-long)
Copied to clipboard
| Challenge: | Large language models (LLMs) exhibit remarkable abilities without being explicitly trained for such tasks. |
| Approach: | They propose a method that unlearns a test dataset and evaluates the unlearned model on training datasets. |
| Outcome: | The proposed method resembles UnTrac, while being efficient for massive training datasets. |
How Pre-trained Language Models Capture Factual Knowledge? A Causal-Inspired Analysis (2022.findings-acl)
Copied to clipboard
Shaobo Li, Xiaoguang Li, Lifeng Shang, Zhenhua Dong, Chengjie Sun, Bingquan Liu, Zhenzhou Ji, Xin Jiang, Qun Liu
| Challenge: | Recent studies show that pre-trained language models can fill in the missing factual words in cloze-style prompts such as ”Dante was born in [MASK]” . |
| Approach: | They propose to quantitatively measure and evaluate the word-level patterns that PLMs depend on to generate the missing factual words. |
| Outcome: | The proposed model fills in the missing factual words in cloze-style prompts by relying on effective clues or shortcut patterns. |
Factual Probing Is [MASK]: Learning vs. Learning to Recall (2021.naacl-main)
Copied to clipboard
| Challenge: | Existing methods for factual probing can interpret the model’s prediction accuracy as a lower bound on the amount of factual information it encodes. |
| Approach: | They propose a method which directly optimizes in continuous embedding space and can predict an additional 6.4% of facts in the LAMA benchmark. |
| Outcome: | The proposed method outperforms the best previous prompt method by 6.4% on the LAMA benchmark. |
Tracing the Roots of Facts in Multilingual Language Models: Independent, Shared, and Transferred Knowledge (2024.eacl-long)
Copied to clipboard
| Challenge: | Using low-resource languages, multilingual language models (ML-LMs) have been developed to transfer factual knowledge across languages. |
| Approach: | They ask how ML-LMs acquire and represent factual knowledge . they use a multilingual factual information probing dataset to investigate ML . |
| Outcome: | The findings highlight the challenge of maintaining consistency factual knowledge across languages. |