WebExpert: domain-aware web agents with critic-guided expert experience for high-precision search
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866910109958930432 |
|---|---|
| author | Hu, Yuelin Cheng, Zhengxue Wu, Ronghua Gu, Qunshan Hu, Hongwei Liu, Wei Liang, Qiao Song, Li |
| author_facet | Hu, Yuelin Cheng, Zhengxue Wu, Ronghua Gu, Qunshan Hu, Hongwei Liu, Wei Liang, Qiao Song, Li |
| contents | Specialized web tasks in finance, biomedicine, and pharmaceuticals remain challenging due to missing domain priors: queries drift, evidence is noisy, and reasoning is brittle. We present WebExpert, a domain-aware web agent that we implement end-to-end, featuring : (i) sentence-level experience retrieval with topic merging and rule distillation, (ii) schemalight facet induction that bootstraps time,region,policy,industry facets from weak supervision instead of static hand-written lexicons, and (iii) preference-optimized planning that jointly improves query planning and retrieval via pairwise preference learning alongside a coverage-aware objective. At inference, a lightweight experience gate biases decoding toward active facets with fallback under low-retrieval confidence. On GAIA, GPQA, HLE, and WebWalkerQA, WebExpert improves Answer Exact Match (EM) by 1.5-3.6 pp over the strongest browsing baseline and reduces page hops. Analysis shows consistent gains and ablations on retrieval, topic merging, facet induction, and preference-aware training. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_06177 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | WebExpert: domain-aware web agents with critic-guided expert experience for high-precision search Hu, Yuelin Cheng, Zhengxue Wu, Ronghua Gu, Qunshan Hu, Hongwei Liu, Wei Liang, Qiao Song, Li Information Retrieval Artificial Intelligence Computation and Language Specialized web tasks in finance, biomedicine, and pharmaceuticals remain challenging due to missing domain priors: queries drift, evidence is noisy, and reasoning is brittle. We present WebExpert, a domain-aware web agent that we implement end-to-end, featuring : (i) sentence-level experience retrieval with topic merging and rule distillation, (ii) schemalight facet induction that bootstraps time,region,policy,industry facets from weak supervision instead of static hand-written lexicons, and (iii) preference-optimized planning that jointly improves query planning and retrieval via pairwise preference learning alongside a coverage-aware objective. At inference, a lightweight experience gate biases decoding toward active facets with fallback under low-retrieval confidence. On GAIA, GPQA, HLE, and WebWalkerQA, WebExpert improves Answer Exact Match (EM) by 1.5-3.6 pp over the strongest browsing baseline and reduces page hops. Analysis shows consistent gains and ablations on retrieval, topic merging, facet induction, and preference-aware training. |
| title | WebExpert: domain-aware web agents with critic-guided expert experience for high-precision search |
| topic | Information Retrieval Artificial Intelligence Computation and Language |
| url | https://arxiv.org/abs/2604.06177 |