Harness-G: A Graph-Structured Harness for Search Agents
- Published
- Source
- arXiv
- Paper number
- 782
- Field
- LLMs / NLP
- arXiv ID
- 2607.27652
Key points
- It finds a retrieval-equivalence collapse phenomenon, where different queries increasingly retrieve the same information as training progresses.
- It replaces free-form query generation with graph-based finite action selection over Select, Lookup, and Answer.
- Structured Non-myopic Credit, or SNC, assigns credit by comparing alternative actions from the same state.
- It improves average F1 by +10.74 points for Graph-R1 at 1.5B and +3.98 points at 3B, reaching the best score on six QA benchmarks.
- Graph construction costs $0 when done programmatically, far cheaper than LLM-based graph construction, which costs $2.81 to $4.14.
Paper links
External research summaries. These are not HDATF publications or measured product results.