# ThinkFL: Self-Refining Failure Localization for Microservice Systems via Reinforcement Fine-Tuning

arXiv: https://arxiv.org/abs/2504.18776 · HTML: https://mlsanity.com/paper/2504.18776/

## Key facts

- Cited by 18 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 26 Apr 2025.

## Built on by (most cited first, 18 of 18)

- [A Survey on Parallel Text Generation: From Parallel Decoding to Diffusion Language Models](https://mlsanity.com/paper/2508.08712/) (2025)
- [A Survey of AIOps in the Era of Large Language Models](https://mlsanity.com/paper/2507.12472/) (2025)
- [Adaptive Root Cause Localization for Microservice Systems with Multi-Agent Recursion-of-Thought](https://mlsanity.com/paper/2508.20370/) (2025)
- [MicroRemed: Benchmarking LLMs in Microservices Remediation](https://mlsanity.com/paper/2511.01166/) (2025)
- [Agentic Memory Enhanced Recursive Reasoning for Root Cause Localization in Microservices](https://mlsanity.com/paper/2601.02732/) (2026)
- [Hypothesize-Then-Verify: Speculative Root Cause Analysis for Microservices with Pathwise Parallelism](https://mlsanity.com/paper/2601.02736/) (2026)
- [LogAction: Consistent Cross-system Anomaly Detection through Logs via Active Domain Adaptation](https://mlsanity.com/paper/2510.03288/) (2025)
- [Efficient Failure Management for Multi-Agent Systems with Reasoning Trace Representation](https://mlsanity.com/paper/2603.21522/) (2026)
- [From Feedback Loops to Policy Updates: Reinforcement Fine-Tuning for LLM-Based Alpha Factor Discovery](https://mlsanity.com/paper/2605.15412/) (2026)
- [RuntimeSlicer: Towards Generalizable Unified Runtime State Representation for Failure Management](https://mlsanity.com/paper/2603.21495/) (2026)
- [OpenRCA 2.0: From Outcome Labels to Causal Process Supervision](https://mlsanity.com/paper/2606.27154/) (2026)
- [R-Log: Incentivizing Log Analysis Capability in LLMs via Reasoning-based Reinforcement Learning](https://mlsanity.com/paper/2509.25987/) (2025)
- [Towards Robust LLM Post-Training: Automatic Failure Management for Reinforcement Fine-Tuning](https://mlsanity.com/paper/2605.04431/) (2026)
- [Towards In-Depth Root Cause Localization for Microservices with Multi-Agent Recursion-of-Thought](https://mlsanity.com/paper/2605.14866/) (2026)
- [JustDiag!: A Diagnostic Justification Engine for Accountable Root Cause Analysis](https://mlsanity.com/paper/2606.19407/) (2026)
- [OpsLLM: Construction of Large Language Model for Software Operations with Multi-stage Learning](https://mlsanity.com/paper/2605.02906/) (2026)
- [Beyond Fault Localization: A Trajectory-Level Study of LLM Agents for Microservice Root Cause Analysis](https://mlsanity.com/paper/2608.21310/) (2026)
- [Playing social deduction games with reinforcement fine-tuned large language models](https://mlsanity.com/paper/2610.04261/) (2026)

## Abstract (arXiv)

As modern microservice systems grow increasingly popular and complex-often consisting of hundreds or even thousands of fine-grained, interdependent components-they are becoming more susceptible to frequent and subtle failures. Ensuring system reliability therefore hinges on accurate and efficient failure localization. Traditional failure localization approaches based on small models lack the flexibility to adapt to diverse failure scenarios, while recent LLM-based methods suffer from two major limitations: they often rely on rigid invocation workflows that constrain the model's ability to dynamically explore optimal localization paths, and they require resource-intensive inference, making them cost-prohibitive for real-world deployment. To address these challenges, we explore the use of reinforcement fine-tuning to equip lightweight LLMs with reasoning and self-refinement capabilities, significantly improving the cost-effectiveness and adaptability of LLM-based failure localization. We begin with an empirical study to identify three key capabilities essential for accurate localization. Building on these insights, we propose a progressive multi-stage GRPO fine-tuning framework, which integrates a multi-factor failure localization grader and a recursion-of-thought actor module. The resulting model, ThinkFL, not only outperforms existing state-of-the-art LLMs and baseline methods in localization accuracy but also reduces end-to-end localization latency from minutes to seconds, demonstrating strong potential for real-world applications.