# Symbiotic Cooperation for Web Agents: Harnessing Complementary Strengths of Large and Small LLMs

arXiv: https://arxiv.org/abs/2502.07942 · HTML: https://mlsanity.com/paper/2502.07942/

## Key facts

- Cited by 24 papers in the map (data through 14 Oct 2026).
- Placed in Web, search, and research agents, one of 16 directions in LLM agents and tool use.
- Builds on 62 of its 101 references that are in the map.
- First posted on arXiv 11 Feb 2025.

## Placed in

- LLM agents and tool use / [Web, search, and research agents](https://mlsanity.com/agents/web-search-and-research-agents/)

## Builds on (most cited first, 30 of 62)

- [Distilling the Knowledge in a Neural Network](https://mlsanity.com/paper/1503.02531/) (2015)
- [DistilBERT, a distilled version of BERT: smaller, faster, cheaper and lighter](https://mlsanity.com/paper/1910.01108/) (2019)
- [SWE-bench: Can Language Models Resolve Real-World GitHub Issues?](https://mlsanity.com/paper/2310.06770/) (2023)
- [Self-Instruct: Aligning Language Models with Self-Generated Instructions](https://mlsanity.com/paper/2212.10560/) (2022)
- [Improving Factuality and Reasoning in Language Models through Multiagent Debate](https://mlsanity.com/paper/2305.14325/) (2023)
- [SWE-agent: Agent-Computer Interfaces Enable Automated Software Engineering](https://mlsanity.com/paper/2405.15793/) (2024)
- [Knowledge Distillation: A Survey](https://mlsanity.com/paper/2006.05525/) (2020)
- [A Survey on LLM-as-a-Judge](https://mlsanity.com/paper/2411.15594/) (2024)
- [WebShop: Towards Scalable Real-World Web Interaction with Grounded Language Agents](https://mlsanity.com/paper/2207.01206/) (2022)
- [Mind2Web: Towards a Generalist Agent for the Web](https://mlsanity.com/paper/2306.06070/) (2023)
- [OSWorld: Benchmarking Multimodal Agents for Open-Ended Tasks in Real Computer Environments](https://mlsanity.com/paper/2404.07972/) (2024)
- [Encouraging Divergent Thinking in Large Language Models through Multi-Agent Debate](https://mlsanity.com/paper/2305.19118/) (2023)
- [Distilling Step-by-Step! Outperforming Larger Language Models with Less Training Data and Smaller Model Sizes](https://mlsanity.com/paper/2305.02301/) (2023)
- [LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory](https://mlsanity.com/paper/2410.10813/) (2024)
- [CogAgent: A Visual Language Model for GUI Agents](https://mlsanity.com/paper/2312.08914/) (2023)
- [Language Agent Tree Search Unifies Reasoning Acting and Planning in Language Models](https://mlsanity.com/paper/2310.04406/) (2023)
- [Improved Knowledge Distillation via Teacher Assistant](https://mlsanity.com/paper/1902.03393/) (2019)
- [A Real-World WebAgent with Planning, Long Context Understanding, and Program Synthesis](https://mlsanity.com/paper/2307.12856/) (2023)
- [Teaching Small Language Models to Reason](https://mlsanity.com/paper/2212.08410/) (2022)
- [Paraphrasing Complex Network: Network Compression via Factor Transfer](https://mlsanity.com/paper/1802.04977/) (2018)
- [FireAct: Toward Language Agent Fine-tuning](https://mlsanity.com/paper/2310.05915/) (2023)
- [WebRL: Training LLM Web Agents via Self-Evolving Online Curriculum Reinforcement Learning](https://mlsanity.com/paper/2411.02337/) (2024)
- [Specializing Smaller Language Models towards Multi-Step Reasoning](https://mlsanity.com/paper/2301.12726/) (2023)
- [Magentic-One: A Generalist Multi-Agent System for Solving Complex Tasks](https://mlsanity.com/paper/2411.04468/) (2024)
- [Chain-of-Thought Reasoning Without Prompting](https://mlsanity.com/paper/2402.10200/) (2024)
- [Agent Q: Advanced Reasoning and Learning for Autonomous AI Agents](https://mlsanity.com/paper/2408.07199/) (2024)
- [Large Language Models for Data Annotation and Synthesis: A Survey](https://mlsanity.com/paper/2402.13446/) (2024)
- [SWE-bench Multimodal: Do AI Systems Generalize to Visual Software Domains?](https://mlsanity.com/paper/2410.03859/) (2024)
- [You Only Look at Screens: Multimodal Chain-of-Action Agents](https://mlsanity.com/paper/2309.11436/) (2023)
- [Mapping Natural Language Instructions to Mobile UI Action Sequences](https://mlsanity.com/paper/2005.03776/) (2020)

## Built on by (most cited first, 24 of 24)

- [A Survey of Context Engineering for Large Language Models](https://mlsanity.com/paper/2507.13334/) (2025)
- [Thinking vs. Doing: Agents that Reason by Scaling Test-Time Interaction](https://mlsanity.com/paper/2506.07976/) (2025)
- [WebChoreArena: Evaluating Web Browsing Agents on Realistic Tedious Web Tasks](https://mlsanity.com/paper/2506.01952/) (2025)
- [Coding Agents with Multimodal Browsing are Generalist Problem Solvers](https://mlsanity.com/paper/2506.03011/) (2025)
- [AnyMAC: Cascading Flexible Multi-Agent Collaboration via Next-Agent Prediction](https://mlsanity.com/paper/2506.17784/) (2025)
- [SynthAgent: Adapting Web Agents with Synthetic Supervision](https://mlsanity.com/paper/2511.06101/) (2025)
- [ToolPRMBench: Evaluating and Advancing Process Reward Models for Tool-using Agents](https://mlsanity.com/paper/2601.12294/) (2026)
- [From Question Answering to Task Completion: A Survey on Agent System and Harness Design](https://mlsanity.com/paper/2606.20683/) (2026)
- [WebOperator: Action-Aware Tree Search for Autonomous Agents in Web Environment](https://mlsanity.com/paper/2512.12692/) (2025)
- [OpAgent: Operator Agent for Web Navigation](https://mlsanity.com/paper/2602.13559/) (2026)
- [WebDART: Dynamic Decomposition and Re-planning for Complex Web Tasks](https://mlsanity.com/paper/2510.06587/) (2025)
- [From Benchmarks to Business Impact: Deploying IBM Generalist Agent in Enterprise Production](https://mlsanity.com/paper/2510.23856/) (2025)
- [Avenir-Web: Human-Experience-Imitating Multimodal Web Agents with Mixture of Grounding Experts](https://mlsanity.com/paper/2602.02468/) (2026)
- [Interaction-Driven Browsing: A Human-in-the-Loop Conceptual Framework Informed by Human Web Browsing for Browser-Using Agents](https://mlsanity.com/paper/2509.12049/) (2025)
- [Co-EPG: A Framework for Co-Evolution of Planning and Grounding in Autonomous GUI Agents](https://mlsanity.com/paper/2511.10705/) (2025)
- [Guided by the Experts: Provable Feature Learning Dynamic of Soft-Routed Mixture-of-Experts](https://mlsanity.com/paper/2510.07205/) (2025)
- [Cognitive Duality for Adaptive Web Agents](https://mlsanity.com/paper/2508.05081/) (2025)
- [ActionEngine: From Reactive to Programmatic GUI Agents via State Machine Memory](https://mlsanity.com/paper/2602.20502/) (2026)
- [Probing to Refine: Reinforcement Distillation of LLMs via Explanatory Inversion](https://mlsanity.com/paper/2603.19266/) (2026)
- [AutoSurfer -- Teaching Web Agents through Comprehensive Surfing, Learning, and Modeling](https://mlsanity.com/paper/2604.27253/) (2026)
- [WebChallenger: A Reliable and Efficient Generalist Web Agent](https://mlsanity.com/paper/2606.10423/) (2026)
- [Software Engineering for and with GUI Agent](https://mlsanity.com/paper/2608.09278/) (2026)
- [Toward Autonomous UI Exploration: The UIExplorer Benchmark](https://mlsanity.com/paper/2506.17779/) (2025)
- [AutoTailor: Automatic, User-Aligned Capability Selection and Adaptation for Web Agents](https://mlsanity.com/paper/2609.13548/) (2026)

## Abstract (arXiv)

Web browsing agents powered by large language models (LLMs) have shown tremendous potential in automating complex web-based tasks. Existing approaches typically rely on large LLMs (e.g., GPT-4o) to explore web environments and generate trajectory data, which is then used either for demonstration retrieval (for large LLMs) or to distill small LLMs (e.g., Llama3) in a process that remains decoupled from the exploration. In this paper, we propose AgentSymbiotic, an iterative framework that couples data synthesis with task-performance, yielding a "symbiotic improvement" for both large and small LLMs. Our study uncovers a complementary dynamic between LLM types: while large LLMs excel at generating high-quality trajectories for distillation, the distilled small LLMs-owing to their distinct reasoning capabilities-often choose actions that diverge from those of their larger counterparts. This divergence drives the exploration of novel trajectories, thereby enriching the synthesized data. However, we also observe that the performance of small LLMs becomes a bottleneck in this iterative enhancement process. To address this, we propose two innovations in LLM distillation: a speculative data synthesis strategy that mitigates off-policy bias, and a multi-task learning approach designed to boost the reasoning capabilities of the student LLM. Furthermore, we introduce a Hybrid Mode for Privacy Preservation to address user privacy concerns. Evaluated on the WEBARENA benchmark, AgentSymbiotic achieves SOTA performance with both LLM types. Our best Large LLM agent reaches 52%, surpassing the previous best of 45%, while our 8B distilled model demonstrates a competitive 49%, exceeding the prior best of 28%. Code will be released upon acceptance.