Guided Search Strategies in Non-Serializable Environments with Applications to Software Engineering Agents

• URL: http://arxiv.org/abs/2505.13652v1
• Date: Mon, 19 May 2025 18:50:15 GMT
• Title: Guided Search Strategies in Non-Serializable Environments with Applications to Software Engineering Agents
• Authors: Karina Zainullina, Alexander Golubev, Maria Trofimova, Sergei Polezhaev, Ibragim Badertdinov, Daria Litvintseva, Simon Karasik, Filipp Fisin, Sergei Skvortsov, Maksim Nekrashevich, Anton Shevtsov, Boris Yangel,
• Abstract summary: Large language models (LLMs) have recently achieved remarkable results in complex multi-step tasks.<n>They often struggle to maintain consistent performance across multiple solution attempts.
• Score: 31.651748374218446
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Large language models (LLMs) have recently achieved remarkable results in complex multi-step tasks, such as mathematical reasoning and agentic software engineering. However, they often struggle to maintain consistent performance across multiple solution attempts. One effective approach to narrow the gap between average-case and best-case performance is guided test-time search, which explores multiple solution paths to identify the most promising one. Unfortunately, effective search techniques (e.g. MCTS) are often unsuitable for non-serializable RL environments, such as Docker containers, where intermediate environment states cannot be easily saved and restored. We investigate two complementary search strategies applicable to such environments: 1-step lookahead and trajectory selection, both guided by a learned action-value function estimator. On the SWE-bench Verified benchmark, a key testbed for agentic software engineering, we find these methods to double the average success rate of a fine-tuned Qwen-72B model, achieving 40.8%, the new state-of-the-art for open-weights models. Additionally, we show that these techniques are transferable to more advanced closed models, yielding similar improvements with GPT-4o.

Related papers

• Training Long-Context, Multi-Turn Software Engineering Agents with Reinforcement Learning [31.540626068273014]
  We train an agent based on Qwen2.5-72B-Instruct to solve real-world software engineering tasks.<n>Our approach increases the agent's success rate on the SWE-bench Verified benchmark from a 20% fine-tuned baseline to 39%.
  arXiv  Detail & Related papers  (2025-08-05T14:30:47Z)
• AgentSwift: Efficient LLM Agent Design via Value-guided Hierarchical Search [58.98450205734779]
  Large language model (LLM) agents have demonstrated strong capabilities across diverse domains.<n>Existing agent search methods suffer from three major limitations.<n>We introduce a comprehensive framework to address these challenges.
  arXiv  Detail & Related papers  (2025-06-06T12:07:23Z)
• SimpleDeepSearcher: Deep Information Seeking via Web-Powered Reasoning Trajectory Synthesis [89.99161034065614]
  Retrieval-augmented generation (RAG) systems have advanced large language models (LLMs) in complex deep search scenarios.<n>Existing approaches face critical limitations that lack high-quality training trajectories and suffer from distributional mismatches.<n>This paper introduces SimpleDeepSearcher, a framework that bridges the gap through strategic data engineering rather than complex training paradigms.
  arXiv  Detail & Related papers  (2025-05-22T16:05:02Z)
• Iterative Tool Usage Exploration for Multimodal Agents via Step-wise Preference Tuning [69.32855772335624]
  Multimodal agents, which integrate a controller e.g., a vision language model, with external tools, have demonstrated remarkable capabilities in tackling complex multimodal tasks.<n>Existing approaches for training these agents depend on extensive human-annotated task-answer pairs and tool trajectories.<n>We propose an iterative tool usage exploration method for multimodal agents without any pre-collected data, namely SPORT.<n>SPORT has four iterative components: task synthesis, step sampling, step verification, and preference tuning.
  arXiv  Detail & Related papers  (2025-04-30T12:01:27Z)
• Thinking Longer, Not Larger: Enhancing Software Engineering Agents via Scaling Test-Time Compute [61.00662702026523]
  We propose a unified Test-Time Compute scaling framework that leverages increased inference-time instead of larger models.<n>Our framework incorporates two complementary strategies: internal TTC and external TTC.<n>We demonstrate our textbf32B model achieves a 46% issue resolution rate, surpassing significantly larger models such as DeepSeek R1 671B and OpenAI o1.
  arXiv  Detail & Related papers  (2025-03-31T07:31:32Z)
• EfficientLLaVA:Generalizable Auto-Pruning for Large Vision-language Models [64.18350535770357]
  We propose an automatic pruning method for large vision-language models to enhance the efficiency of multimodal reasoning.<n>Our approach only leverages a small number of samples to search for the desired pruning policy.<n>We conduct extensive experiments on the ScienceQA, Vizwiz, MM-vet, and LLaVA-Bench datasets for the task of visual question answering.
  arXiv  Detail & Related papers  (2025-03-19T16:07:04Z)
• POGEMA: A Benchmark Platform for Cooperative Multi-Agent Pathfinding [76.67608003501479]
  We introduce POGEMA, a comprehensive set of tools that includes a fast environment for learning, a problem instance generator, and a visualization toolkit.<n>We also introduce and define an evaluation protocol that specifies a range of domain-related metrics, computed based on primary evaluation indicators.<n>The results of this comparison, which involves a variety of state-of-the-art MARL, search-based, and hybrid methods, are presented.
  arXiv  Detail & Related papers  (2024-07-20T16:37:21Z)
• Efficient Multi-agent Reinforcement Learning by Planning [33.51282615335009]
  Multi-agent reinforcement learning (MARL) algorithms have accomplished remarkable breakthroughs in solving large-scale decision-making tasks. Most existing MARL algorithms are model-free, limiting sample efficiency and hindering their applicability in more challenging scenarios. We propose the MAZero algorithm, which combines a centralized model with Monte Carlo Tree Search (MCTS) for policy search.
  arXiv  Detail & Related papers  (2024-05-20T04:36:02Z)
• Maximize to Explore: One Objective Function Fusing Estimation, Planning, and Exploration [87.53543137162488]
  We propose an easy-to-implement online reinforcement learning (online RL) framework called textttMEX. textttMEX integrates estimation and planning components while balancing exploration exploitation automatically. It can outperform baselines by a stable margin in various MuJoCo environments with sparse rewards.
  arXiv  Detail & Related papers  (2023-05-29T17:25:26Z)
• Heuristic-free Optimization of Force-Controlled Robot Search Strategies in Stochastic Environments [13.622757453459748]
  Even relatively simple peg-in-hole tasks are typically subject to variations, requiring search motions to find relevant features such as holes. This paper introduces an automatic, data-driven and conditioning-free approach to optimize search strategies. We evaluate our approach on two different industrial robots in the context of spiral and probe search for THT electronics assembly.
  arXiv  Detail & Related papers  (2022-07-15T15:16:08Z)
• MetaDistiller: Network Self-Boosting via Meta-Learned Top-Down Distillation [153.56211546576978]
  In this work, we propose that better soft targets with higher compatibil-ity can be generated by using a label generator. We can employ the meta-learning technique to optimize this label generator. The experiments are conducted on two standard classificationbenchmarks, namely CIFAR-100 and ILSVRC2012.
  arXiv  Detail & Related papers  (2020-08-27T13:04:27Z)
• Fast and stable MAP-Elites in noisy domains using deep grids [1.827510863075184]
  Deep-Grid MAP-Elites is a variant of the MAP-Elites algorithm that uses an archive of similar previously encountered solutions to approximate the performance of a solution. We show that this simple approach is significantly more resilient to noise on the behavioural descriptors, while achieving competitive performances in terms of fitness optimisation.
  arXiv  Detail & Related papers  (2020-06-25T08:47:23Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.