Research map / Large language models / Language, safety and interpretability
One Homonym per Translation
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 |
|---|---|---|---|
| Evaluate whether each translation maps to only one homonym | English homonyms in MSC · MSC (Italian) | support (%): 99.9 | Helped |
| Evaluate whether each translation maps to only one homonym | English homonyms in JSC · JSC (Japanese) | support (%): 99.8 | Helped |
| Restrict the one-sense-per-discourse expectation to homonyms | Homonymous words in SemCor · SemCor documents | support (%): 99.6 | Helped |
| Restrict the one-sense-per-collocation expectation to homonyms | IMS trained on SemCor · Five benchmark datasets | effective accuracy: 97.9% | Helped |
| Require sense clusters not to combine senses from distinct homonyms | OntoNotes clustering · OntoNotes sense clusters | support (%): 99.9 | Helped |
| Use homonym pairs rather than OntoNotes sense-cluster pairs | Translation partitioning in MSC · MSC, 20 pairs of each type | pairs with strict translation partitioning: 16 of the 20 vs. 6 of the 20; p < 0.005 | Helped |
Closest papers by meaning
- Monitoring Latent World States in Language Models with Propositional Probes (ICLR 2025 · similarity 0.84)
- DOCS: Quantifying Weight Similarity for Deeper Insights into Large Language Models (ICLR 2025 · similarity 0.84)
- From Tokens to Words: On the Inner Lexicon of LLMs (ICLR 2025 · similarity 0.83)
- Translate Meanings, Not Just Words: IdiomKB’s Role in Optimizing Idiomatic Translation with Language Models (AAAI 2024 · similarity 0.83)
- The Same but Different: Structural Similarities and Differences in Multilingual Language Modeling (ICLR 2025 · similarity 0.83)
- The Geometry of Categorical and Hierarchical Concepts in Large Language Models (ICLR 2025 · similarity 0.83)
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.