The authors designed CSEG, a six-layer decision tree that returns an ordered label from 1 to 7 based on source independence, conflict detection, and adjudication by source authority and recency.
Grounding rewards must check source independence and conflict handling in search agents
The authors introduce a six-layer decision hierarchy, a benchmark of 1,160 verified trajectories, and compact reward models that improve judged grounding quality.
Chinese Tech
Haochen Zhao · Minsheng Hao · Zheng Li · Yifan Li · Zikang Liu · Bowen Qin · +5 more
Peking University · Tencent · Renmin University of China
Research Digest··2 min read
Zhao et al.
Why this paper
From Tencent and 2 others
In one line
A hierarchical reward for source independence and conflict handling improves search-agent evidence grounding beyond outcome and citation checks.
What we could check
- ·No code link found
- ·No weights link found
- ·No dataset link found
- ✓Compute or model size stated (params 4B)
- ·No stated limitations found
- ✓Reports numbers on named benchmarks (3 benchmarks)
Observed from the paper text and links we have. Absence here means we did not find it, not that it does not exist.
§