Question Answering on RULER-HQA 14K context length
0.8125Normalized Sub-EM AccuracyLycheeMemory-7B w/o Gate
Evaluation results
| Method | Links | ||
|---|---|---|---|
| LycheeMemory-7B w/o GateModel=LycheeMemory-7B, Gate=disabled2026.02 | 0.8125 | — | |
| RL-MemAgent-7BModel=RL-MemAgent-7B2026.02 | 0.7969 | — | |
| LycheeMemory-7BModel=LycheeMemory-7B, Gate=enabled2026.02 | 0.7656 | 1.2 | |
| QwenLong-L1-32BModel=QwenLong-L1-32B2026.02 | 0.75 | — | |
| Search-R1Model=Search-R12026.02 | 0.7188 | — | |
| DS-Distill-Qwen-32BModel=DS-Distill-Qwen-32B2026.02 | 0.6641 | — | |
| RAG + Qwen2.5-7B-InstructModel=RAG + Qwen2.5-7B-Instruct2026.02 | 0.6641 | — | |
| DS-Distill-Qwen-14BModel=DS-Distill-Qwen-14B2026.02 | 0.6484 | — | |
| Qwen2.5-Instruct-14B-1MModel=Qwen2.5-Instruct-14B-1M2026.02 | 0.6094 | — | |
| Qwen2.5-Instruct-7B-1MModel=Qwen2.5-Instruct-7B-1M2026.02 | 0.5625 | — | |
| DS-Distill-Qwen-7BModel=DS-Distill-Qwen-7B2026.02 | 0.125 | — |