ARCO introduces a co-evolving rubric model with generation and scoring heads plus a trajectory decomposition constraint that improves exact-match scores on multi-hop QA tasks over outcome, rubric, and process reward baselines.
Reinforcement learning for long-horizon multi-turn search agents
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
ARCO: Adaptive Rubric with Co-Evolution for Multi-Step LLM-Based Agents
ARCO introduces a co-evolving rubric model with generation and scoring heads plus a trajectory decomposition constraint that improves exact-match scores on multi-hop QA tasks over outcome, rubric, and process reward baselines.