Word Sense Induction on SemEval Task 13 2013
23.7FNMIPolyLM large
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| PolyLM largeSize=Large2026.03 | 23.7 | — | — | 66.7 | |
| PolyLM baseSize=Base2026.03 | 23 | — | — | 65.4 | |
| bertwsidynamic patterns=true2019.05 | 21.4 | 64 | 37 | — | |
| LSDPSubstitutes=BERT-large2026.03 | 21.1 | — | — | 64.1 | |
| bertwsidynamic patterns=false2019.05 | 19.3 | 63.6 | 35.1 | — | |
| GPT-4oMethodology=LLM-based WSI prompting2026.03 | 16.9 | — | — | 58.6 | |
| LSDPnumber of clusters=72019.05 | 11.3 | 57.5 | 25.4 | — | |
| Llama 3.3 70BMethodology=LLM-based WSI prompting, Parameters=70B2026.03 | 8.9 | — | — | 44.2 | |
| AutoSense2019.05 | 7.96 | 61.7 | 22.16 | — | |
| MCC-S2019.05 | 7.62 | 55.6 | 20.58 | — | |
| ST(SW)embedding similarity weighting=true2019.05 | 7.14 | 55.4 | 19.89 | — | |
| 1cpexStrategy=One cluster per example2026.03 | 6.9 | — | — | — | |
| AI-KU2019.05 | 6.5 | 39 | 15.92 | — | |
| Llama 3.1 8BMethodology=LLM-based WSI prompting, Parameters=8B2026.03 | 2.3 | — | — | 57.1 | |
| 1cplStrategy=One cluster per lemma2026.03 | 0 | — | — | 61.23 |