Research map / Large language models / Language, safety and interpretability
Enhancing Natural Language Inference Using New and Expanded Training Data Sets and New Learning Models
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| add NER-CHANGED data to SNLI training | ESIM · NER-CHANGED test | test accuracy (%): 56.11 -> 69.77 | Helped |
| add NER-CHANGED data to SNLI training | DecAtt · NER-CHANGED test | test accuracy (%): 33.80 -> 82.91 | Helped |
| add ROLE-SWITCHED data to SNLI training | ESIM · ROLE-SWITCHED test | test accuracy (%): 50.46 -> 93.72 | Helped |
| add ROLE-SWITCHED data to SNLI training | BERT · ROLE-SWITCHED test | test accuracy (%): 49.74 -> 49.78 | No clear effect |
| add symbolic-similarity lambda attention | ESIM · NER-CHANGED test (SNLI+NC train) | test accuracy (%): 69.77 -> 99.13 | Helped |
| add symbolic-similarity lambda attention | DecAtt · NER-CHANGED test (SNLI+NC train) | test accuracy (%): 82.91 -> 97.81 | Helped |
| add symbolic-similarity lambda attention | ESIM · SNLI test (SNLI+NC train) | test accuracy (%): 85.09 -> 87.10 | Helped |
| add symbolic-similarity lambda attention | DecAtt · SNLI test (SNLI+NC train) | test accuracy (%): 83.50 -> 82.94 | No clear effect |
| add symbolic-similarity lambda attention | ESIM · ROLE-SWITCHED test (SNLI+RS+NC train) | test accuracy (%): 84.45 -> 87.92 | Mixed |
| add symbolic-similarity lambda attention | DecAtt · ROLE-SWITCHED test (SNLI+RS+NC train) | test accuracy (%): 50.11 -> 50.08 | No clear effect |
| add symbolic-similarity lambda attention | ESIM · NER-CHANGED test (SNLI only train) | test accuracy (%): 56.11 -> 53.95 | No clear effect |
| add symbolic-similarity lambda attention | ESIM · NER-CHANGED test (MNLI+NC train) | test accuracy (%): 75.58 -> 95.12 | Helped |
Closest papers by meaning
- Improving Natural Language Inference Using External Knowledge in the Science Questions Domain (AAAI 2019 · similarity 0.89)
- Supervising Model Attention with Human Explanations for Robust Natural Language Inference (AAAI 2022 · similarity 0.89)
- Improving Natural Language Understanding for LLMs via Large-Scale Instruction Synthesis (AAAI 2025 · similarity 0.85)
- Generative Judge for Evaluating Alignment (ICLR 2024 · similarity 0.85)
- SeqGPT: An Out-of-the-Box Large Language Model for Open Domain Sequence Understanding (AAAI 2024 · similarity 0.84)
- Supervised Knowledge Makes Large Language Models Better In-context Learners (ICLR 2024 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.