Agent Alpha: Tree Search Unifying Generation, Exploration and Evaluation for Computer-Use Agents

arXiv 2602.02995 ↗First posted 3 Feb 2026cs.AIcited by 12

On the map: LLM reasoning → Reasoning for agents, planning and code

also in 1 more

LLM agents and tool use → General LLM agent frameworks and benchmarks

Abstract (arXiv)

While scaling test-time compute through trajectory-level sampling has significantly improved Graphical User Interface (GUI) agents, the lack of regressive ability prevents the reuse of partial successes and the recovery from early missteps. In this paper, we introduce Agent Alpha, a unified framework that synergizes generation, exploration, and evaluation through step-level Monte Carlo Tree Search (MCTS). It enables active modeling or exploiting structures of the planning space. By integrating alpha-UCT guided search into the interaction loop, Agent Alpha enables deliberate planning, facilitating early pruning of suboptimal branches and efficient prefix reuse. We also employ comparison-driven evaluation to mitigate absolute scoring biases and diversity-constrained expansion to maintain a compact, informative search space. Regret bound of alpha-UCT is analyzed. On the OSWorld benchmark, Agent Alpha achieves a state-of-the-art success rate of ~ 77%, significantly outperforming trajectory-level baselines under equivalent compute.

Key facts. Cited by 12 papers in the map (data through 7 Oct 2026). Placed in Reasoning for agents, planning and code, one of 17 directions in LLM reasoning. Builds on 33 of its 44 references that are in the map. First posted on arXiv 3 Feb 2026.

Builds on

33 of 44 references.

Closest ancestors · 1

Same direction, newest first

2025Rethinking Testing for LLM Applications: Characteristics, Challenges, and a Lightweight Interaction Protocol28 Aug

Other references · 32

2022Chain-of-Thought Prompting Elicits Reasoning in Large Language Modelscited by 12,302
2022ReAct: Synergizing Reasoning and Acting in Language Modelscited by 8,355
2023Let's Verify Step by Stepcited by 3,598
2023Tree of Thoughts: Deliberate Problem Solving with Large Language Modelscited by 3,155
2024Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameterscited by 1,928
2024A Survey on LLM-as-a-Judgecited by 1,384
2025s1: Simple test-time scalingcited by 1,311
2024OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environmentscited by 1,134
2023Kosmos-2: Grounding Multimodal Large Language Models to the Worldcited by 1,075
2017Mastering Chess and Shogi by Self-Play with a General Reinforcement Learning Algorithmcited by 723
2025UI-TARS: Pioneering Automated GUI Interaction with Native Agentscited by 584
2024Executable Code Actions Elicit Better LLM Agentscited by 537

20 more

2024AndroidWorld: A Dynamic Benchmarking Environment for Autonomous Agentscited by 450
2024Trial and Error: Exploration-Based Trajectory Optimization for LLM Agentscited by 194
2025GTA1: GUI Test-time Scaling Agentcited by 102
2024ScreenAgent: A Vision Language Model-driven Computer Control Agentcited by 101
2025Scaling Test-time Compute for LLM Agentscited by 59
2025CoAct-1: Computer-using Multi-Agent System with Coding Actionscited by 33
2025Assessing Consistency and Reproducibility in the Outputs of Large Language Models: Evidence Across Diverse Finance and Accounting Taskscited by 19
2026OS-Symphony: A Holistic Framework for Robust and Generalist Computer-Using Agentcited by 19
2024Modeling Other Players with Bayesian Beliefs for Games with Incomplete Informationcited by 16
2025Scaling Agents for Computer Usecited by 14
2025InSPO: Unlocking Intrinsic Self-Reflection for LLM Preference Optimizationcited by 11
2025Optimizing Prompt Sequences using Monte Carlo Tree Search for LLM-Based Optimizationcited by 9
2024Learning from Random Demonstrations: Offline Reinforcement Learning with Importance-Sampled Diffusion Modelscited by 8
2025Lipschitz Lifelong Monte Carlo Tree Search for Mastering Non-Stationary Taskscited by 8
2025A Neurosymbolic Framework for Interpretable Cognitive Attack Detection in Augmented Realitycited by 8
2022Coordinate-Aligned Multi-Camera Collaboration for Active Multi-Object Trackingcited by 7
2025MALinZero: Efficient Low-Dimensional Search for Mastering Complex Multi-Agent Planningcited by 5
2024Mixed Text Recognition with Efficient Parameter Fine-Tuning and Transformercited by 4
2025Tail-Risk-Safe Monte Carlo Tree Search under PAC-Level Guaranteescited by 4
2025Perception Graph for Cognitive Attack Reasoning in Augmented Realitycited by 4

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 12.

12 papers.

2026From Question Answering to Task Completion: A Survey on Agent System and Harness Designcited by 9
2026RoboRouter: Training-Free Policy Routing for Robotic Manipulationcited by 6
2026NonZero: Interaction-Guided Exploration for Multi-Agent Monte Carlo Tree Searchcited by 4
2026Operator-Guided Invariance Learning for Continuous Reinforcement Learningcited by 4
2026Metric-Gradient Projection for Stable Multi-Agent Policy Learningcited by 3
2026Efficient Agent Evaluation via Diversity-Guided User Simulationcited by 2
2026Interactive Critique-Revision Training for Reliable Structured LLM Generationcited by 2
2026IntentScore: Intent-Conditioned Action Evaluation for Computer-Use Agentscited by 1
2026Software Engineering for and with GUI Agentcited by 1
2026TClone: Low-Latency Forking of Live GUI Environments for Computer-Use Agentscited by 0
2026VisCritic: Visual State Comparison as Process Reward for GUI Agentscited by 0
2026FlowEdit: Information-Theoretic Control of LLM Reasoning Flows for Ill-posed Problems Involving Conflictscited by 0

Similar, not cited

Close in wording, no citation either way