2608.24667

Total: 1

#1 EviGraph: Towards Verifiable Evidence Construction for Information-Seeking Agents [PDF1] [Copy] [Kimi] [REL]

Authors: Jiashun Chen, Yirong Mao, Wenhui Que

Agentic Web search can retrieve relevant information without establishing that the retrieved content actually supports the claims used in an answer. Existing agents typically keep search and evidence recording in a linear interaction trace and optimize primarily for final-answer correctness, providing limited supervision for intermediate grounding. We present EviGraph, a deep-search framework that separates search execution from evidence recording while using a shared policy for the trainable roles. An executor plans concise queries, a frozen evidence verifier inspects source pages and returns verbatim evidence items with an explicit polarity, and the policy maps those items to add/support graph requests that are checked by a deterministic structural validator. The resulting graph serves both as persistent working memory and as a source of dense process rewards, enabling reinforcement learning to directly supervise evidence construction rather than only the final answer. On BrowseComp-Plus, a Qwen3-8B EviGraph agent achieves 35.9% accuracy under a matched interaction budget, compared with 26.9% for the same dual-role architecture without reinforcement learning and 2.7% for a monolithic agent, while generating fewer tokens per rollout. Consistent gains on BrowseComp, GAIA, and XBench indicate that explicitly structuring and rewarding evidence recording improves agentic search

Subject: Information Retrieval

Publish: 2026-08-25 15:06:54 UTC