Web Navigation on WebShop Explicit Structural Drift II
95Success Rate (Source)Llama3.3-70B + GLOVE
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| Llama3.3-70B + GLOVEBackbone=Llama3.3-70B, GLOVE Augmentation=true2026.01 | 95 | 88.8 | 81.3 | |
| Qwen3-30B + GLOVEBackbone=Qwen3-30B, GLOVE Augmentation=true2026.01 | 95 | 86.2 | 78.7 | |
| GPT-4o + GLOVEBackbone=GPT-4o, GLOVE Augmentation=true2026.01 | 90 | 92.5 | 87.5 | |
| Grok-3 + GLOVEBackbone=Grok-3, GLOVE Augmentation=true2026.01 | 87.5 | 93.8 | 86.3 | |
| DeepSeek-V3.2 + GLOVEBackbone=DeepSeek-V3.2, GLOVE Augmentation=true2026.01 | 85 | 93.8 | 87.6 | |
| Llama3.1-8B + GLOVEBackbone=Llama3.1-8B, GLOVE Augmentation=true2026.01 | 67.5 | 75 | 70 | |
| Qwen2.5-7B + GLOVEBackbone=Qwen2.5-7B, GLOVE Augmentation=true2026.01 | 63.7 | 71.2 | 67.4 | |
| Llama3.3-70BBackbone=Llama3.3-70B, GLOVE Augmentation=false2026.01 | 8.8 | 7.5 | — | |
| Qwen3-30BBackbone=Qwen3-30B, GLOVE Augmentation=false2026.01 | 8.8 | 7.5 | — | |
| Grok-3Backbone=Grok-3, GLOVE Augmentation=false2026.01 | 8.8 | 7.5 | — | |
| Llama3.1-8BBackbone=Llama3.1-8B, GLOVE Augmentation=false2026.01 | 7.5 | 5 | — | |
| Qwen2.5-7BBackbone=Qwen2.5-7B, GLOVE Augmentation=false2026.01 | 7.5 | 3.8 | — | |
| GPT-4oBackbone=GPT-4o, GLOVE Augmentation=false2026.01 | 5 | 5 | — | |
| DeepSeek-V3.2Backbone=DeepSeek-V3.2, GLOVE Augmentation=false2026.01 | 5 | 6.2 | — |