Retrieval-Augmented Question Answering on DeepSearch HotpotQA
56Success RateActiveContext
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ActiveContextTaskExecutor=Gemini-3.0-flash2026.04 | 56 | 7.4 | |
| Full ContextTaskExecutor=Gemini-3.0-flash2026.04 | 52 | 53.5 | |
| Training-Free GRPOTaskExecutor=Gemini-3.0-flash2026.04 | 46 | 60.5 | |
| ActiveContextTaskExecutor=gpt-4o2026.04 | 38 | 7.4 | |
| Gemini-3.0-flashTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 38 | 8.7 | |
| Training-Free GRPOTaskExecutor=gpt-4o2026.04 | 34 | 35.5 | |
| ContextCuratorTaskExecutor=gpt-4o-mini, Category=RL-Trained Curator2026.04 | 34 | 7.3 | |
| SynapseTaskExecutor=Gemini-3.0-flash2026.04 | 32 | 149.2 | |
| Full ContextTaskExecutor=Gemini-2.5-flash2026.04 | 32 | 35.5 | |
| Full ContextTaskExecutor=gpt-4o2026.04 | 32 | 35.5 | |
| Gemini-2.5-flashTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 32 | 8.3 | |
| ActiveContextTaskExecutor=Gemini-2.5-flash2026.04 | 31.8 | 8.4 | |
| Mem0TaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 31 | 52.1 | |
| Full ContextTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 30 | 37.2 | |
| Training-Free GRPOTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 30 | 39.1 | |
| gpt-4oTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 30 | 11.4 | |
| Training-Free GRPOTaskExecutor=Gemini-2.5-flash2026.04 | 28 | 76.5 | |
| gpt-4o-miniTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 28 | 10.9 | |
| Qwen2.5-7B-InstructTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 28 | 10 | |
| SynapseTaskExecutor=Gemini-2.5-flash2026.04 | 26 | 135.3 | |
| AWMTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 25 | 36.2 | |
| SynapseTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 20 | 55.9 | |
| SynapseTaskExecutor=gpt-4o2026.04 | 18 | 52.1 |