Topic Modeling on TweetTopic
0.392CvERNIE-4.5-0.3B
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ERNIE-4.5-0.3BMethod Identity=Ours, Soft Label Source SLM=ERNIE-4.5-0.3B2026.02 | 0.392 | 2.9 | 0.989 | 0.781 | |
| Llama-3.1-8BMethod Identity=Ours, Soft Label Source SLM=Llama-3.1-8B2026.02 | 0.384 | 2.88 | 0.993 | 0.774 | |
| Llama-3.2-1BMethod Identity=Ours, Soft Label Source SLM=Llama-3.2-1B2026.02 | 0.383 | 2.9 | 0.993 | 0.781 | |
| BERTopicContextual Embedding Backbone=GTE-large-en-v1.52026.02 | 0.364 | 2.19 | 0.996 | 0.562 | |
| CombinedTMContextual Embedding Backbone=GTE-large-en-v1.52026.02 | 0.36 | 2.36 | 0.988 | 0.588 | |
| ZeroshotTMContextual Embedding Backbone=GTE-large-en-v1.52026.02 | 0.356 | 2.3 | 0.994 | 0.573 | |
| ECRTM2026.02 | 0.356 | 1.85 | 1 | 0.399 | |
| ProdLDA2026.02 | 0.355 | 2.11 | 0.994 | 0.533 | |
| ETM2026.02 | 0.351 | 2.22 | 0.936 | 0.552 | |
| LDA2026.02 | 0.35 | 1.95 | 0.992 | 0.441 | |
| FASTopicContextual Embedding Backbone=GTE-large-en-v1.52026.02 | 0.276 | 1.96 | 0.626 | 0.557 |