arxiv arXiv cs.LG · just now · research

GRASP framework improves strategic planning with multi-stage agentic AI

from English

Researchers introduce GRASP, a strategy-aware, multi-stage planning framework designed to generate high-quality natural language executable plans for complex tasks by decoupling the pipeline into specialized modules. The system pre-compiles global macro-guidelines via GenPlan, explores localized strategies through RevPlan, and evaluates trajectories using a multi-criteria discriminator called VerPlan.

GRASP establishes a new state-of-the-art frontier across diverse datasets, yielding accuracy gains of ~12.4% on Natural Plan Calendar Scheduling and ~30.8% on ZebraLogic over direct LLM planners.
The framework flattens the multi-task degradation penalty seen in standard planners, achieving an absolute accuracy gain of up to 16.7% in interleaved dual-task environments.
By isolating context and enforcing strict macro-regularization, GRASP outperforms frontier reasoning models like GPT-5-mini by a margin of 14.5%.

GRASP addresses the reliability degradation of LLMs as task complexity increases, offering a robust solution for complex planning scenarios where traditional methods fail.

Importance 2/3 arXiv cs.LG AI agents Research paper

Related articles

arxiv arXiv cs.AI · 1h ago · 1 view Live

GRASP introduces strategy-aware multi-stage planning to improve LLM reliability

Researchers introduce GRASP, a strategy-aware, multi-stage planning framework designed to generate high-quality natural language executable plans for complex tasks where standard Large Language Models typically degrade in reliability. The system decouples the planning pipeline into specialized modules: GenPlan for global macro-guidelines, RevPlan for exploring localized strategies, and VerPlan for independent trajectory evaluation.

lab Anthropic News · 2d ago · 29 views

Anthropic's Claude discovers novel ART enzyme system in bacteriophages

Anthropic has announced a new life sciences research group dedicated to using Claude to identify uncharacterized protein families and test hypotheses through laboratory experiments. In early results, the AI autonomously discovered a novel enzyme system called array-associated reverse transcriptases (ART) found in bacteriophages.

arxiv arXiv cs.AI · 2d ago · 16 views

AIDE^2 enables recursive self-improvement of AI research agents

Researchers present AIDE^2, a system that implements a loop of recursive self-improvement for an AI research agent by optimizing its own code. The system proposes changes to its internal logic, benchmarks modified versions on AI R&D tasks, and retains only those improvements that perform best on hidden evaluations.

arxiv arXiv cs.AI · 2d ago · 17 views

Growing Harness turns recurring agent control into reusable code via failure-guided training

The authors introduce Growing Harness, a training paradigm that learns an agent's control structure from task feedback rather than relying on standard context-heavy harnesses. By converting recurring control decisions into executable code and reserving LLM calls for semantic reasoning, the method aims to create reusable specialist agents.

arxiv arXiv cs.LG · 2d ago · 16 views

AIDE^2 enables recursive self-improvement of AI research agents

Researchers present AIDE^2, a system that implements a loop of recursive self-improvement for a frontier AI research agent by optimizing its own code. The system proposes changes to its internal logic, benchmarks modified versions on AI R&D tasks, and retains only those updates that perform best on hidden evaluations.