Retrieval-Augmented Question Answering on DeepSearch TriviaQA
78Success Rate (SR)ActiveContext
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ActiveContextTaskExecutor=Gemini-3.0-flash2026.04 | 78 | 3.2 | |
| Full ContextTaskExecutor=Gemini-3.0-flash2026.04 | 76 | 20.4 | |
| Training-Free GRPOTaskExecutor=Gemini-3.0-flash2026.04 | 76 | 16.7 | |
| SynapseTaskExecutor=Gemini-3.0-flash2026.04 | 68 | 51.2 | |
| ActiveContextTaskExecutor=Gemini-2.5-flash2026.04 | 66 | 3.7 | |
| ActiveContextTaskExecutor=gpt-4o2026.04 | 66 | 3.6 | |
| Gemini-3.0-flashTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 66 | 4.4 | |
| Training-Free GRPOTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 64 | 18.3 | |
| Gemini-2.5-flashTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 64 | 4.6 | |
| Mem0TaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 62 | 27.6 | |
| ContextCuratorTaskExecutor=gpt-4o-mini, Category=RL-Trained Curator2026.04 | 62 | 5 | |
| Training-Free GRPOTaskExecutor=Gemini-2.5-flash2026.04 | 60 | 38.4 | |
| Full ContextTaskExecutor=Gemini-2.5-flash2026.04 | 58 | 20.6 | |
| Full ContextTaskExecutor=gpt-4o2026.04 | 58 | 20.7 | |
| Training-Free GRPOTaskExecutor=gpt-4o2026.04 | 58 | 19.6 | |
| gpt-4oTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 58 | 6.9 | |
| AWMTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 56 | 16.4 | |
| gpt-4o-miniTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 56 | 7.3 | |
| Qwen2.5-7B-InstructTaskExecutor=gpt-4o-mini, Category=Zero-Shot Curators2026.04 | 52 | 6.5 | |
| Full ContextTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 50 | 19.7 | |
| SynapseTaskExecutor=gpt-4o-mini, Category=Baselines & Retrieval-Based Memory2026.04 | 37 | 42.7 | |
| SynapseTaskExecutor=gpt-4o2026.04 | 36 | 42 | |
| SynapseTaskExecutor=Gemini-2.5-flash2026.04 | 35 | 86.1 |