# Agent Task on FrozenLake Source: [Agent Task on FrozenLake]() Collection index: [llms.txt]() Data updated: 2026-02-26T06:04:54.588Z Evaluation of agent performance on the FrozenLake environment, measured by Success Rate (SR) and Average Utility Value (AUV). A research result applies to its stated dataset, split, metric and evaluation configuration. Compare like-for-like settings, preserve the metric's unit and direction, and distinguish paper-reported results from independent reproduction. The best result recorded here is not a claim of exhaustive, current global coverage. | Scope | Recorded value | | --- | --- | | Task | Agent Task | | Dataset (including any stated split) | FrozenLake | | Metric | Success Rate | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 17 of 17 | | Coverage | All results currently returned for this benchmark | Evaluation settings may differ between results. The original papers and recorded configurations are needed to establish comparability. A shared metric name alone does not establish an identical protocol. ## Metrics - [Success Rate](): Higher is better; unit: PERCENT. - [Average Utility Value](): Higher is better; unit: SCALAR. ## Recorded results | Method | Success Rate | Result date | Report paper | | --- | --- | --- | --- | | Gemini 2.5 Pro | 100 | Not recorded | 698181aca8b01e76cc8e27c2 | | gpt-oss-120b | 99 | Not recorded | 698181aca8b01e76cc8e27c2 | | DeepSeek-V3.2 | 97 | Not recorded | 698181aca8b01e76cc8e27c2 | | Gemini 2.5 Flash | 97 | Not recorded | 698181aca8b01e76cc8e27c2 | | DeepSeek-R1 | 96 | Not recorded | 698181aca8b01e76cc8e27c2 | | Qwen3-30B-A3B-Thinking | 93 | Not recorded | 698181aca8b01e76cc8e27c2 | | Qwen3-4B-Thinking | 91 | Not recorded | 698181aca8b01e76cc8e27c2 | | GLM-4-32B-0414 | 67 | Not recorded | 698181aca8b01e76cc8e27c2 | | Phi-4 | 66 | Not recorded | 698181aca8b01e76cc8e27c2 | | Phi-4-reasoning | 62 | Not recorded | 698181aca8b01e76cc8e27c2 | | Qwen3-30B-A3B-Instruct | 61 | Not recorded | 698181aca8b01e76cc8e27c2 | | Qwen3-4B-Instruct | 59 | Not recorded | 698181aca8b01e76cc8e27c2 | | Ministral-3-14B-Instruct | 56 | Not recorded | 698181aca8b01e76cc8e27c2 | | Llama-3.3-70B-Instruct | 44 | Not recorded | 698181aca8b01e76cc8e27c2 | | Llama-3.1-8B-Instruct | 12 | Not recorded | 698181aca8b01e76cc8e27c2 | | GLM-4-9B-Chat | 10 | Not recorded | 698181aca8b01e76cc8e27c2 | | Mistral-7B-Instruct | 6 | Not recorded | 698181aca8b01e76cc8e27c2 | ### Gemini 2.5 Pro Success Rate: 100. Result ID: 69960b235698665ec444057e. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### gpt-oss-120b Success Rate: 99. Result ID: 69960b235698665ec444057c. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### DeepSeek-V3.2 Success Rate: 97. Result ID: 69960b235698665ec4440577. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Gemini 2.5 Flash Success Rate: 97. Result ID: 69960b235698665ec4440578. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### DeepSeek-R1 Success Rate: 96. Result ID: 69960b235698665ec444057d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### Qwen3-30B-A3B-Thinking Success Rate: 93. Result ID: 69960b235698665ec444057a. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### Qwen3-4B-Thinking Success Rate: 91. Result ID: 69960b235698665ec4440579. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### GLM-4-32B-0414 Success Rate: 67. Result ID: 69960b235698665ec4440573. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Phi-4 Success Rate: 66. Result ID: 69960b235698665ec4440576. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Phi-4-reasoning Success Rate: 62. Result ID: 69960b235698665ec444057b. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Thinking Model"}. This record does not assert independent reproduction. ### Qwen3-30B-A3B-Instruct Success Rate: 61. Result ID: 69960b235698665ec444056f. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Qwen3-4B-Instruct Success Rate: 59. Result ID: 69960b235698665ec444056e. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Ministral-3-14B-Instruct Success Rate: 56. Result ID: 69960b235698665ec4440575. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Llama-3.3-70B-Instruct Success Rate: 44. Result ID: 69960b235698665ec4440571. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Llama-3.1-8B-Instruct Success Rate: 12. Result ID: 69960b235698665ec4440570. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### GLM-4-9B-Chat Success Rate: 10. Result ID: 69960b235698665ec4440572. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. ### Mistral-7B-Instruct Success Rate: 6. Result ID: 69960b235698665ec4440574. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Non-thinking Model"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.