KARL: Knowledge Agents via Reinforcement Learning

TL;DR

KARL introduces a reinforcement learning-based system for training enterprise search agents that excel across diverse tasks, leveraging a new evaluation suite, synthetic data generation, and multi-task training for superior performance.

Contribution

The paper presents KARL, a novel reinforcement learning framework with a new evaluation suite, synthetic data generation pipeline, and a sample-efficient training paradigm for enterprise search agents.

Findings

KARL outperforms Claude 4.6 and GPT 5.2 on KARLBench.

Models trained on heterogeneous tasks generalize better.

KARL surpasses strong closed models with sufficient compute.

Abstract

We present a system for training enterprise search agents via reinforcement learning that achieves state-of-the-art performance across a diverse suite of hard-to-verify agentic search tasks. Our work makes four core contributions. First, we introduce KARLBench, a multi-capability evaluation suite spanning six distinct search regimes, including constraint-driven entity search, cross-document report synthesis, tabular numerical reasoning, exhaustive entity retrieval, procedural reasoning over technical documentation, and fact aggregation over internal enterprise notes. Second, we show that models trained across heterogeneous search behaviors generalize substantially better than those optimized for any single benchmark. Third, we develop an agentic synthesis pipeline that employs long-horizon reasoning and tool use to generate diverse, grounded, and high-quality training data, with iterative bootstrapping from increasingly capable models. Fourth, we propose a new post-training paradigm based on iterative large-batch off-policy RL that is sample efficient, robust to train-inference engine discrepancies, and naturally extends to multi-task training with out-of-distribution generalization. Compared to Claude 4.6 and GPT 5.2, KARL is Pareto-optimal on KARLBench across cost-quality and latency-quality trade-offs, including tasks that were out-of-distribution during training. With sufficient test-time compute, it surpasses the strongest closed models. These results show that tailored synthetic data in combination with multi-task reinforcement learning enables cost-efficient and high-performing knowledge agents for grounded reasoning.