Context Adherence on Representative guardrail dataset
96F1-ScoreChainPoll
Evaluation results
| Method | Links | |
|---|---|---|
| ChainPollModel=GPT 4.12026.02 | 96 | |
| Luna-2Model=Llama 3.2 3B2026.02 | 95 | |
| Single tokenModel=Llama 3.2 3B2026.02 | 43 |
| Method | Links | |
|---|---|---|
| ChainPollModel=GPT 4.12026.02 | 96 | |
| Luna-2Model=Llama 3.2 3B2026.02 | 95 | |
| Single tokenModel=Llama 3.2 3B2026.02 | 43 |