mlsanityData to 7 Oct 2026

When Does Meaning Backfire? Investigating the Role of AMRs in NLI

arXiv 2506.14613 ↗First posted 17 Jun 2025cs.CLcited by 0

Not on the map. Why?

Abstract (arXiv)

Natural Language Inference (NLI) relies heavily on adequately parsing the semantic content of the premise and hypothesis. In this work, we investigate whether adding semantic information in the form of an Abstract Meaning Representation (AMR) helps pretrained language models better generalize in NLI. Our experiments integrating AMR into NLI in both fine-tuning and prompting settings show that the presence of AMR in fine-tuning hinders model generalization while prompting with AMR leads to slight gains in GPT-4o. However, an ablation study reveals that the improvement comes from amplifying surface-level differences rather than aiding semantic reasoning. This amplification can mislead models to predict non-entailment even when the core meaning is preserved.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 28 of its 45 references that are in the map. First posted on arXiv 17 Jun 2025.

Builds on

28 of 45 references.

References · 28

  1. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  2. 2019RoBERTa: A Robustly Optimized BERT Pretraining Approachcited by 13,266
  3. 2019BART: Denoising Sequence-to-Sequence Pre-training for Natural Language Generation, Translation, and Comprehensioncited by 5,874
  4. 2018GLUE: A Multi-Task Benchmark and Analysis Platform for Natural Language Understandingcited by 4,854
  5. 2017A Broad-Coverage Challenge Corpus for Sentence Understanding through Inferencecited by 2,665
  6. 2015A large annotated corpus for learning natural language inferencecited by 1,837
  7. 2019Right for the Wrong Reasons: Diagnosing Syntactic Heuristics in Natural Language Inferencecited by 864
  8. 2019Is BERT Really Robust? A Strong Baseline for Natural Language Attack on Text Classification and Entailmentcited by 663
  9. 2018Annotation Artifacts in Natural Language Inference Datacited by 656
  10. 2021Do Prompt-Based Models Really Understand the Meaning of their Prompts?cited by 324
  11. 2018Hypothesis Only Baselines in Natural Language Inferencecited by 291
  12. 2019Don't Take the Easy Way Out: Ensemble Based Methods for Avoiding Known Dataset Biasescited by 279
16 more
  1. 2023Can ChatGPT Understand Too? A Comparative Study on ChatGPT and Fine-tuned BERTcited by 155
  2. 2019BERTs of a feather do not generalize together: Large variability in generalization across models with similar test set performancecited by 89
  3. 2020Syntactic Data Augmentation Increases Robustness to Inference Heuristicscited by 85
  4. 2019Human vs. Muppet: A Conservative Estimate of Human Performance on the GLUE Benchmarkcited by 32
  5. 2021Perturbing Inputs for Fragile Interpretations in Deep Natural Language Processingcited by 30
  6. 2021NeuralLog: Natural Language Inference with Joint Neural and Logical Reasoningcited by 24
  7. 2020The Curse of Performance Instability in Analysis Datasets: Consequences, Source, and Suggestionscited by 22
  8. 2023SMATCH++: Standardized and Extended Evaluation of Semantic Graphscited by 19
  9. 2025Instruction-Tuning LLMs for Event Extraction with Annotation Guidelinescited by 18
  10. 2021Translate, then Parse! A strong baseline for Cross-Lingual AMR Parsingcited by 15
  11. 2024Analyzing the Role of Semantic Representations in the Era of Large Language Modelscited by 12
  12. 2023Lost in Translationese? Reducing Translation Effect Using Abstract Meaning Representationcited by 10
  13. 2018Acquisition of Phrase Correspondences using Natural Deduction Proofscited by 8
  14. 2020Learning as Abduction: Trainable Natural Logic Theorem Prover for Natural Language Inferencecited by 6
  15. 2025Revisiting Prompt Optimization with Large Reasoning Models-A Case Study on Event Extractioncited by 6
  16. 2024Semantic Graphs for Syntactic Simplification: A Revisit from the Age of LLMcited by 4

Built on by

Not cited yet.