Evolve on Graph CWQ - Hit@1: leaderboard
Evolve on Graph v1 evaluates knowledge-graph reasoning using direct answers, chain of thought, self-consistency, ReAct, retrieval, or separately trained Solver/Proposer systems. It uses Qwen2.5-3B/7B-Instruct and Llama3.1-8B-Instruct. Training lasts 140 steps, with 5 rollouts per task and at most 15 assistant turns. Hit@1 requires an exact normalized match to any gold answer; F1 measures overlap between predicted and gold answer lists. Freebase, Wikidata, and task-specific KG-R1 training scopes remain distinct. Trained weights are not aliased to the instruction backbone. Metric: Hit@1 (%).
Metric: Hit@1 (%). Source: arxiv.org. Saturation forecast: Rough model projection: around 2026. 23 models tracked.
Top models
| # | Model | Score | Overall rank |
|---|---|---|---|
| 1 | EoG study Qwen-2.5-7B-Instruct KG-R1 trainedCWQ | 58.03 | |
| 2 | EoG study Qwen-2.5-7B-Instruct KG-R1 trainedWebQSP | 50.92 | |
| 3 | EoG study Qwen-2.5-7B-Instruct EoG | 50.58 | |
| 4 | EoG study LLaMA-3.1-8B-Instruct EoG | 50.36 | |
| 5 | EoG study Qwen-2.5-7B-Instruct KG-R1 trainedGrailQA | 46.02 | |
| 6 | EoG study LLaMA-3.1-8B-Instruct SC | 43.39 | |
| 7 | EoG study Qwen-2.5-7B-Instruct SC | 38.97 | |
| 8 | EoG study Qwen-2.5-3B-Instruct EoG | 38.14 | |
| 9 | EoG study Qwen-2.5-7B-Instruct Graph Reranker | 37.69 | |
| 10 | EoG study Qwen-2.5-7B-Instruct ToG | 36.9 |
Interactive version: theaggregate.ai/benchmark?slug=evolve-on-graph-cwq-hit-1 · How It Works · Data refreshed daily, snapshot 2026-10-28.