# Infusing Finetuning with Semantic Dependencies

arXiv: https://arxiv.org/abs/2012.05395 · HTML: https://mlsanity.com/paper/2012.05395/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 10 Dec 2020.

## Built on by (most cited first, 18 of 18)

- [Recent Advances in Natural Language Processing via Large Pre-Trained Language Models: A Survey](https://mlsanity.com/paper/2111.01243/) (2021)
- [Implicit Representations of Meaning in Neural Language Models](https://mlsanity.com/paper/2106.00737/) (2021)
- [The Semantic Hub Hypothesis: Language Models Share Semantic Representations Across Languages and Modalities](https://mlsanity.com/paper/2411.04986/) (2024)
- [Language Models Use Monotonicity to Assess NPI Licensing](https://mlsanity.com/paper/2105.13818/) (2021)
- [How Does Fine-tuning Affect the Geometry of Embedding Space: A Case Study on Isotropy](https://mlsanity.com/paper/2109.04740/) (2021)
- [Micromodels for Efficient, Explainable, and Reusable Systems: A Case Study on Mental Health](https://mlsanity.com/paper/2109.13770/) (2021)
- [Linguistic Frameworks Go Toe-to-Toe at Neuro-Symbolic Language Modeling](https://mlsanity.com/paper/2112.07874/) (2021)
- [Bird's Eye: Probing for Linguistic Graph Structures with a Simple Information-Theoretic Approach](https://mlsanity.com/paper/2105.02629/) (2021)
- [Transparency Helps Reveal When Language Models Learn Meaning](https://mlsanity.com/paper/2210.07468/) (2022)
- [Schr"odinger's Tree -- On Syntax and Neural Language Models](https://mlsanity.com/paper/2110.08887/) (2021)
- [Learning with Latent Structures in Natural Language Processing: A Survey](https://mlsanity.com/paper/2201.00490/) (2022)
- [Ask to Understand: Question Generation for Multi-hop Question Answering](https://mlsanity.com/paper/2203.09073/) (2022)
- [An Empirical Revisiting of Linguistic Knowledge Fusion in Language Understanding Tasks](https://mlsanity.com/paper/2210.13002/) (2022)
- [Mixture-of-Linguistic-Experts Adapters for Improving and Interpreting Pre-trained Language Models](https://mlsanity.com/paper/2310.16240/) (2023)
- [Assessing Logical Reasoning Capabilities of Encoder-Only Transformer Models](https://mlsanity.com/paper/2312.11720/) (2023)
- [The Grammar of Transformers: A Systematic Review of Interpretability Research on Syntactic Knowledge in Language Models](https://mlsanity.com/paper/2601.19926/) (2026)
- [Rethinking Relation Classification with Graph Meaning Representations](https://mlsanity.com/paper/2310.09772/) (2023)
- [Exploring Graph Representations of Logical Forms for Language Modeling](https://mlsanity.com/paper/2505.14523/) (2025)

## Abstract (arXiv)

For natural language processing systems, two kinds of evidence support the use of text representations from neural language models "pretrained" on large unannotated corpora: performance on application-inspired benchmarks (Peters et al., 2018, inter alia), and the emergence of syntactic abstractions in those representations (Tenney et al., 2019, inter alia). On the other hand, the lack of grounded supervision calls into question how well these representations can ever capture meaning (Bender and Koller, 2020). We apply novel probes to recent language models -- specifically focusing on predicate-argument structure as operationalized by semantic dependencies (Ivanova et al., 2012) -- and find that, unlike syntax, semantics is not brought to the surface by today's pretrained models. We then use convolutional graph encoders to explicitly incorporate semantic parses into task-specific finetuning, yielding benefits to natural language understanding (NLU) tasks in the GLUE benchmark. This approach demonstrates the potential for general-purpose (rather than task-specific) linguistic supervision, above and beyond conventional pretraining and finetuning. Several diagnostics help to localize the benefits of our approach.