Reasoning on Countries
57F1 ScoreSkyline
Evaluation results
| Method | Links | ||
|---|---|---|---|
| SkylineSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 57 | 57 | |
| KVCommSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B, Communication Ratio (M)=0.52025.10 | 57 | 57 | |
| KVCommSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B, Communication Ratio (M)=0.72025.10 | 57 | 57 | |
| SkylineSender Model (Ms)=ehristoforu/falcon3-ultraset, Receiver Model (Mr)=huihui-ai/Falcon3-7B-Instruct-abliterated2025.10 | 56 | 56 | |
| SkylineSender Model (Ms)=Orion-zhen/Qwen2.5-7B-Instruct-Uncensored, Receiver Model (Mr)=bespokelabs/Bespoke-Stratos-7B2025.10 | 51 | 51 | |
| KVCommSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B, Communication Ratio (M)=0.32025.10 | 46 | 46 | |
| NLDSender Model (Ms)=ehristoforu/falcon3-ultraset, Receiver Model (Mr)=huihui-ai/Falcon3-7B-Instruct-abliterated2025.10 | 46 | 46 | |
| KVCommSender Model (Ms)=ehristoforu/falcon3-ultraset, Receiver Model (Mr)=huihui-ai/Falcon3-7B-Instruct-abliterated, Communication Ratio (M)=0.32025.10 | 46 | 46 | |
| NLDSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 43 | 43 | |
| CIPHERSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 42 | 42 | |
| KVCommSender Model (Ms)=Orion-zhen/Qwen2.5-7B-Instruct-Uncensored, Receiver Model (Mr)=bespokelabs/Bespoke-Stratos-7B, Communication Ratio (M)=0.72025.10 | 41 | 41 | |
| NLDSender Model (Ms)=Orion-zhen/Qwen2.5-7B-Instruct-Uncensored, Receiver Model (Mr)=bespokelabs/Bespoke-Stratos-7B2025.10 | 21 | 21 | |
| BaselineSender Model (Ms)=ehristoforu/falcon3-ultraset, Receiver Model (Mr)=huihui-ai/Falcon3-7B-Instruct-abliterated2025.10 | 8 | 8 | |
| BaselineSender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 5 | 5 | |
| CIPHERSender Model (Ms)=Orion-zhen/Qwen2.5-7B-Instruct-Uncensored, Receiver Model (Mr)=bespokelabs/Bespoke-Stratos-7B2025.10 | 4 | 4 | |
| AC (mean)Sender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 3 | 3 | |
| AC (sum)Sender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 2 | 2 | |
| BaselineSender Model (Ms)=Orion-zhen/Qwen2.5-7B-Instruct-Uncensored, Receiver Model (Mr)=bespokelabs/Bespoke-Stratos-7B2025.10 | 1 | 1 | |
| AC (replace)Sender Model (Ms)=huihui-ai/Llama-3.2-3B-Instruct-abliterated, Receiver Model (Mr)=suayptalha/DeepSeek-R1-Distill-Llama-3B2025.10 | 0 | 0 |