Web Navigation on WebWalkerQA
35Internalized InferenceExperience Internalization
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 12026.06 | 35 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 32026.06 | 34.6 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 12026.06 | 33.9 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 32026.06 | 33.1 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 12026.06 | 32.9 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 22026.06 | 32.4 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 32026.06 | 32.2 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 22026.06 | 31.8 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 32026.06 | 31.5 | — | |
| Experience InternalizationBase Model=Qwen3-8B-Instruct, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=On-policy distillation, Iteration=iter 22026.06 | 31.5 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 22026.06 | 31 | 14.1 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 22026.06 | 30.7 | 34.6 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 12026.06 | 30.6 | 31.5 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 32026.06 | 30 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 12026.06 | 29 | 32.1 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=On-policy distillation, Iteration=iter 12026.06 | 29 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Step-wise injection, Distillation Regime=Off-policy distillation, Iteration=iter 22026.06 | 28.5 | 31.3 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 12026.06 | 25.9 | 28.1 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=On-policy distillation, Iteration=iter 22026.06 | 22.5 | — | |
| Qwen3-8B-InstructExperience Status=w/o experience2026.06 | 21.8 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 12026.06 | 21 | 19.9 | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=On-policy distillation, Iteration=iter 32026.06 | 19.9 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 22026.06 | 18.7 | 9 | |
| Qwen3-4B-Instruct-2507Experience Status=w/o experience2026.06 | 16.6 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=DeepSeek-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 32026.06 | 12.8 | — | |
| Experience InternalizationBase Model=Qwen3-4B-Instruct-2507, Experience Source=Qwen-generated, Injection Pattern=Global injection, Distillation Regime=Off-policy distillation, Iteration=iter 32026.06 | 8.5 | — | |
| Qwen3-4B-Instruct-2507Experience Source=Qwen-generated experience2026.06 | — | 18.5 | |
| Qwen3-4B-Instruct-2507Experience Source=DeepSeek-generated experience2026.06 | — | 25.9 | |
| Qwen3-8B-InstructExperience Source=DeepSeek-generated experience2026.06 | — | 27.79 |