Web Agent Task Success on WebArena Multi 29 tasks
20.7Task Success Rate (SR)ReasoningBank + MaTTS
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| ReasoningBank + MaTTSBackbone LLM=Gemini-2.5-pro2025.09 | 20.7 | 7.2 | |
| ReasoningBank + MaTTSBackbone LLM=Gemini-2.5-flash2025.09 | 17.2 | 8 | |
| ReasoningBankBackbone LLM=Gemini-2.5-flash2025.09 | 13.8 | 8.8 | |
| ReasoningBankBackbone LLM=Gemini-2.5-pro2025.09 | 13.8 | 8.2 | |
| No MemoryBackbone LLM=Gemini-2.5-flash2025.09 | 10.3 | 10 | |
| SynapseBackbone LLM=Gemini-2.5-flash2025.09 | 10.3 | 10.5 | |
| ReasoningBank + MaTTSBackbone LLM=Claude-3.7-sonnet2025.09 | 10.3 | 9.1 | |
| No MemoryBackbone LLM=Gemini-2.5-pro2025.09 | 6.9 | 8.8 | |
| SynapseBackbone LLM=Gemini-2.5-pro2025.09 | 6.9 | 9 | |
| AWMBackbone LLM=Gemini-2.5-flash2025.09 | 3.4 | 7.7 | |
| AWMBackbone LLM=Gemini-2.5-pro2025.09 | 3.4 | 9.3 | |
| ReasoningBankBackbone LLM=Claude-3.7-sonnet2025.09 | 3.4 | 10.5 | |
| No MemoryBackbone LLM=Claude-3.7-sonnet2025.09 | 0 | 11.6 | |
| SynapseBackbone LLM=Claude-3.7-sonnet2025.09 | 0 | 11.8 | |
| AWMBackbone LLM=Claude-3.7-sonnet2025.09 | 0 | 12.4 |