Sentiment Analysis on Financial PhraseBank (FPB)
88.54Macro F1 ScoreSC
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| SCModel=Mistral, Shots=8-shot2025.05 | 88.54 | — | — | |
| Base LLMModel=Mistral, Shots=8-shot2025.05 | 85.8 | — | — | |
| SCModel=Llama, Shots=8-shot2025.05 | 85.58 | — | — | |
| CCModel=Mistral, Shots=8-shot2025.05 | 84.64 | — | — | |
| DCModel=Mistral, Shots=8-shot2025.05 | 83.59 | — | — | |
| BCModel=Mistral, Shots=8-shot2025.05 | 83.59 | — | — | |
| BCModel=Llama, Shots=8-shot2025.05 | 82.71 | — | — | |
| DCModel=Llama, Shots=8-shot2025.05 | 82.44 | — | — | |
| CCModel=Llama, Shots=8-shot2025.05 | 80.97 | — | — | |
| Base LLMModel=Llama, Shots=8-shot2025.05 | 80.26 | — | — | |
| SCModel=Qwen, Shots=8-shot2025.05 | 74.97 | — | — | |
| BCModel=Qwen, Shots=8-shot2025.05 | 66.87 | — | — | |
| CCModel=Qwen, Shots=8-shot2025.05 | 35.96 | — | — | |
| DCModel=Qwen, Shots=8-shot2025.05 | 35.92 | — | — | |
| Base LLMModel=Qwen, Shots=8-shot2025.05 | 30.13 | — | — | |
| BERT2023.10 | — | 83.3 | 81.73 | |
| FiLM# Tokens (Financial Corpus)=2.4B2023.10 | — | 86.25 | 84.48 | |
| FiLM (5.5B)# Tokens (Financial Corpus)=5.5B2023.10 | — | 86.14 | 84.11 | |
| FinBERT-A# Tokens (Financial Corpus)=237M2023.10 | — | 85.25 | 82.45 | |
| FinBERT-Y# Tokens (Financial Corpus)=4.9B2023.10 | — | 83.68 | 82.52 | |
| FLANG-BERT# Tokens (Financial Corpus)=NA2023.10 | — | 84.76 | 83.12 | |
| FLANG-ROBERTa# Tokens (Financial Corpus)=NA2023.10 | — | 83.86 | 82.19 | |
| ROBERTa2023.10 | — | 85.3 | 83.93 | |
| SEC-BERT# Tokens (Financial Corpus)=3.1B2023.10 | — | 84.37 | 82.18 |