Search papers, labs, and topics across Lattice.
This paper introduces PRO-LONG, a programmatic memory framework designed to enhance long-horizon reasoning in large language model (LLM) agents by efficiently managing context through a structured interaction log. The approach addresses the critical tradeoff in context management, where preserving more information can hinder the retrieval of relevant details, thereby improving performance on continual learning benchmarks. PRO-LONG achieves an average improvement of 18.0 percentage points on the ARC-AGI-3 public game set compared to baseline coding agents, while also using significantly fewer tokens than existing specialized harnesses.
PRO-LONG boosts LLM performance on long-horizon tasks by 18% while using up to 5.8 times fewer tokens than traditional methods.
Long-horizon tasks require sustained perception, reasoning, and exploration, and are a persistent challenge for large language model (LLM) agents. This gap is reflected in their limited performance on continual learning benchmarks such as ARC-AGI-3, especially when models are evaluated out of the box. Various agent harnesses have been proposed to close this gap, and each commits to a strategy for handling long sequences of observations, i.e., what information to save from the environment and how to load it into model context, a choice we argue is particularly consequential. Existing methods for context management face a significant tradeoff, as preserving more information makes retrieving relevant details less tractable. We propose PRO-LONG, a minimal context management framework built around programmatic memory for LLM agents in long-horizon, exploratory settings. PRO-LONG addresses the tradeoff by keeping a complete, structured interaction log and capitalizing on recent progress in coding agents to search this history efficiently. On the full ARC-AGI-3 public game set, PRO-LONG improves over a base coding agent by an average of 18.0 percentage points across frontier models, and matches or exceeds state-of-the-art specialized harnesses (up to 76.1% pass@1) while using 4.2-5.8x fewer tokens. With Fable 5, PRO-LONG achieves 97.4% best@2 at a total cost of \$1,750. Relevant code and logs are available at https://github.com/alexisfox7/PRO-LONG.
