General Evaluation on Aggregate Benchmarks
93.9Average ScoreGemini 2.5 Pro+ASP
Evaluation Results
| Method | Links | |
|---|---|---|
| Gemini 2.5 Pro+ASPInference Pipeline=LLM+ASP, Prompt Reference Type=Compact Reference2026.04 | 93.9 | |
| Gemini 2.5 Flash+ASPInference Pipeline=LLM+ASP, Prompt Reference Type=Compact Reference2026.04 | 86.9 | |
| o4-mini+ASPInference Pipeline=LLM+ASP, Prompt Reference Type=Compact Reference2026.04 | 82.3 | |
| DS-R1-0528+ASPInference Pipeline=LLM+ASP, Prompt Reference Type=Compact Reference2026.04 | 81.9 | |
| GOLFModel Backbone=Qwen-3-8B2026.03 | 69.26 | |
| Rubric-as-RewardModel Backbone=Qwen-3-8B2026.03 | 67.08 | |
| Pairwise-GRPOModel Backbone=Qwen-3-8B2026.03 | 66.97 | |
| Critique-GRPOModel Backbone=Qwen-3-8B2026.03 | 66.96 | |
| o4-miniInference Pipeline=Baseline2026.04 | 63.4 | |
| Direct-LikertModel Backbone=Qwen-3-8B2026.03 | 62.99 | |
| Gemini 2.5 ProInference Pipeline=Baseline2026.04 | 62.9 | |
| Qwen-3-8BModel Backbone=Qwen-3-8B2026.03 | 53.95 | |
| DS-R1-0528Inference Pipeline=Baseline2026.04 | 52.3 | |
| δ-MemBackbone=Qwen3-4B-Instruct, Memory Configuration=TSW2026.05 | 51.66 | |
| δ-MemBackbone=Qwen3-4B-Instruct, Memory Configuration=SSW2026.05 | 51.44 | |
| δ-MemBackbone=Qwen3-8B, Memory Configuration=SSW2026.05 | 50.86 | |
| δ-MemBackbone=Qwen3-8B, Memory Configuration=MSW2026.05 | 50.84 | |
| δ-MemBackbone=Qwen3-4B-Instruct, Memory Configuration=MSW2026.05 | 50.74 | |
| δ-MemBackbone=Qwen3-8B, Memory Configuration=TSW2026.05 | 50.68 | |
| GOLFModel Backbone=Llama-3.1-8B-Instruct2026.03 | 50.19 | |
| Qwen3-8BBackbone=Qwen3-8B2026.05 | 47.2 | |
| Qwen3-4B-InstructBackbone=Qwen3-4B-Instruct2026.05 | 46.79 | |
| Context2LoRABackbone=Qwen3-8B2026.05 | 45.92 | |
| DS-V3+ASPInference Pipeline=LLM+ASP, Prompt Reference Type=Compact Reference2026.04 | 45 | |
| Context2LoRABackbone=Qwen3-4B-Instruct2026.05 | 44.9 | |
| Gemini 2.5 FlashInference Pipeline=Baseline2026.04 | 43.3 | |
| Critique-GRPOModel Backbone=Llama-3.1-8B-Instruct2026.03 | 40.92 | |
| Rubric-as-RewardModel Backbone=Llama-3.1-8B-Instruct2026.03 | 40.11 | |
| Pairwise-GRPOModel Backbone=Llama-3.1-8B-Instruct2026.03 | 39.94 | |
| δ-MemBackbone=SmolLM3-3B, Memory Configuration=MSW2026.05 | 36.96 | |
| δ-MemBackbone=SmolLM3-3B, Memory Configuration=SSW2026.05 | 36.67 | |
| Direct-LikertModel Backbone=Llama-3.1-8B-Instruct2026.03 | 35.79 | |
| Context2LoRABackbone=SmolLM3-3B2026.05 | 34.94 | |
| δ-MemBackbone=SmolLM3-3B, Memory Configuration=TSW2026.05 | 34.74 | |
| DS-V3Inference Pipeline=Baseline2026.04 | 34.3 | |
| SmolLM3-3BBackbone=SmolLM3-3B2026.05 | 26.08 | |
| Llama-3.1-8B-InstructModel Backbone=Llama-3.1-8B-Instruct2026.03 | 24.3 |