Truthfulness Evaluation on TruthfulQA
16.9Reliability ScoreAligner
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| AlignerAligner size=13B, Upstream LLM=Alpaca-7B, Zero-shot=true2024.02 | 16.9 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Beaver-7B, Zero-shot=true2024.02 | 13 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Alpaca-7B, Zero-shot=true2024.02 | 11.8 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Llama2-70B-Chat, Zero-shot=true2024.02 | 10.8 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Beaver-7B, Zero-shot=true2024.02 | 10.3 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Llama2-70B-Chat, Zero-shot=true2024.02 | 10.3 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Alpaca-7B, Zero-shot=true2024.02 | 10 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Llama2-70B-Chat, Zero-shot=true2024.02 | 9.1 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Vicuna-13B, Zero-shot=true2024.02 | 7.6 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Vicuna-13B, Zero-shot=true2024.02 | 7.1 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Vicuna-7B, Zero-shot=true2024.02 | 6.6 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Vicuna-33B, Zero-shot=true2024.02 | 6.1 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Beaver-7B, Zero-shot=true2024.02 | 5.4 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Vicuna-33B, Zero-shot=true2024.02 | 5.1 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Vicuna-7B, Zero-shot=true2024.02 | 4.9 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Llama2-7B-Chat, Zero-shot=true2024.02 | 4.2 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Llama2-7B-Chat, Zero-shot=true2024.02 | 3.9 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Vicuna-33B, Zero-shot=true2024.02 | 3.2 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Vicuna-7B, Zero-shot=true2024.02 | 2.7 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Vicuna-13B, Zero-shot=true2024.02 | 2.7 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Claude 2, Zero-shot=true2024.02 | 2 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Llama2-13B-Chat, Zero-shot=true2024.02 | 1.7 | — | — | |
| AlignerAligner size=7B, Upstream LLM=Llama2-7B-Chat, Zero-shot=true2024.02 | 1.5 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Llama2-13B-Chat, Zero-shot=true2024.02 | 1.2 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Llama2-13B-Chat, Zero-shot=true2024.02 | 1 | — | — | |
| AlignerAligner size=2B, Upstream LLM=GPT-3.5, Zero-shot=true2024.02 | 0.7 | — | — | |
| AlignerAligner size=2B, Upstream LLM=Claude 2, Zero-shot=true2024.02 | 0.7 | — | — | |
| AlignerAligner size=13B, Upstream LLM=GPT-3.5, Zero-shot=true2024.02 | 0.5 | — | — | |
| AlignerAligner size=13B, Upstream LLM=Claude 2, Zero-shot=true2024.02 | 0.5 | — | — | |
| AlignerAligner size=7B, Upstream LLM=GPT-3.5, Zero-shot=true2024.02 | 0 | — | — | |
| AlignerAligner size=13B, Upstream LLM=GPT-4, Zero-shot=true2024.02 | -0.2 | — | — | |
| AlignerAligner size=2B, Upstream LLM=GPT-4, Zero-shot=true2024.02 | -0.5 | — | — | |
| AlignerAligner size=7B, Upstream LLM=GPT-4, Zero-shot=true2024.02 | -1.2 | — | — | |
| CBS2025.05 | — | 64 | 98 | |
| Falcon InstructParameter Count=7B2024.02 | — | 27.2 | — | |
| Gemma-2BParameters=2B2024.07 | — | 44 | — | |
| GIFTModel Size=7B2025.10 | — | 69.05 | — | |
| GIFTModel Size=32B2025.10 | — | 70.02 | — | |
| GRPOModel Size=7B2025.10 | — | 66.93 | — | |
| GRPOModel Size=32B2025.10 | — | 69.09 | — | |
| InstructModel Size=7B2025.10 | — | 64.79 | — | |
| InstructModel Size=32B2025.10 | — | 65.58 | — | |
| Llama-2-ChatParameter Count=7B2024.02 | — | 26.3 | — | |
| Mistral-7BParameters=7B2024.07 | — | 41 | — | |
| MPT ChatParameter Count=7B2024.02 | — | 42.7 | — | |
| OLMo (base)Parameter Count=7B2024.02 | — | 31.6 | — | |
| OLMo+SFTParameter Count=7B2024.02 | — | 41.2 | — | |
| OLMo+SFT+DPOParameter Count=7B2024.02 | — | 52 | — | |
| Phi-2Parameters=2.7B2024.07 | — | 42 | — | |
| Phi-2 + BiasDPOParameters=2.7B, Alignment=BiasDPO2024.07 | — | 45 | — | |
| Re-Control2025.05 | — | 74 | 99 | |
| RPJ-INCITE ChatParameter Count=7B2024.02 | — | 53 | — | |
| SEA2025.05 | — | 76 | 99 | |
| SFT2025.05 | — | 62 | 100 | |
| StableLM-3BParameters=3B2024.07 | — | 36 | — | |
| TÜLU 2Parameter Count=7B2024.02 | — | 51.7 | — |