Document Ranking on MS MARCO Document 200 sampled queries (test)
93Top-5 Success RateCORE (Reasoning)
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| CORE (Reasoning)LLM=Claude-4, Optimization Strategy=Reasoning2026.02 | 93 | 88.5 | 81 | |
| CORE (Reasoning)LLM=GPT-4o, Optimization Strategy=Reasoning2026.02 | 91 | 86 | 80 | |
| CORE (Review)LLM=Gemini-2.5, Optimization Strategy=Review2026.02 | 90.5 | 85.5 | 78 | |
| CORE (Review)LLM=Claude-4, Optimization Strategy=Review2026.02 | 89.5 | 85 | 78 | |
| CORE (Review)LLM=GPT-4o, Optimization Strategy=Review2026.02 | 88 | 83 | 76 | |
| CORE (Reasoning)LLM=Gemini-2.5, Optimization Strategy=Reasoning2026.02 | 88 | 83 | 76.5 | |
| CORE (Reasoning)LLM=Grok-3, Optimization Strategy=Reasoning2026.02 | 86.5 | 82 | 75 | |
| CORE (Review)LLM=Grok-3, Optimization Strategy=Review2026.02 | 86 | 81 | 74.5 | |
| CORE (String)LLM=Claude-4, Optimization Strategy=String2026.02 | 61.5 | 50 | 34 | |
| CORE (String)LLM=GPT-4o, Optimization Strategy=String2026.02 | 58.5 | 48.5 | 33.5 | |
| CORE (String)LLM=Gemini-2.5, Optimization Strategy=String2026.02 | 56 | 46 | 31 | |
| CORE (String)LLM=Grok-3, Optimization Strategy=String2026.02 | 54 | 44.5 | 30.5 |