Research map / Large language models / Language, safety and interpretability
Writing Polishment with Simile: Task, Dataset and A Neural Approach
Jiayi Zhang, Zhi Cui, Xiaoqiang Xia, Yalong Guo, Yanran Li, Chen Wei, Jianwei Cui
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| Use beam search with beam size 20 instead of greedy decoding (ablation) | Locate&Gen · CS test | BLEU1: 39.73 -> 41.28 | Improved |
| Use beam search with beam size 20 instead of greedy decoding (ablation) | Locate&Gen · CS test | Dist-S: 0.887 -> 0.855 | Hurt |
| Use beam search with beam size 20 instead of greedy decoding (ablation) | Locate&Gen · CS test | generation length: 7.9 -> 7.2 | Hurt |
| Use pretrained BERT instead of a randomly initialized encoder (ablation) | Locate&Gen · CS test | BLEU1: 36.36 -> 39.73 | Improved |
| Use pretrained BERT instead of a randomly initialized encoder (ablation) | Locate&Gen · CS test | positioning accuracy: 0.713 -> 0.769 | Improved |
| Add insertion bias to the decoder (ablation) | Locate&Gen · CS test | BLEU1: 36.98 -> 39.73 | Improved |
| Add insertion bias to the decoder (ablation) | Locate&Gen · CS test | positioning accuracy: 0.726 -> 0.769 | Improved |
| Use retrieval rather than generation for simile selection | BM25, MatchCBOW, and MatchBERT · CS test | coherence: Locate&Gen 0.77; retrieval methods 0.54, 0.43, 0.64 | Hurt |
Closest papers by meaning
- ScImage: How good are multimodal large language models at scientific text-to-image generation? (ICLR 2025 · similarity 0.82)
- AI as Humanity’s Salieri: Quantifying Linguistic Creativity of Language Models via Systematic Attribution of Machine Text against Web Text (ICLR 2025 · similarity 0.81)
- Bayesian Test-Time Inference of Task-Aligned Similarity from Weak Interactive Feedback (NeurIPS 2026 · similarity 0.81)
- The Alignment Illusion in Multimodal Large Language Models (NeurIPS 2026 · similarity 0.81)
- On Affine Homotopy between Language Encoders (NeurIPS 2024 · similarity 0.81)
- OpenING: A Comprehensive Benchmark for Judging Open-ended Interleaved Image-Text Generation (CVPR 2025 · similarity 0.81)
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.
