# Online Auditing on AFTRAJ-2K Source: [Online Auditing on AFTRAJ-2K]() Collection index: [llms.txt]() Data updated: 2026-05-12T09:30:54.251Z Evaluation of online auditing performance on the AFTRAJ-2K dataset, measured using Exact-F1 and ASS scores across Math, Coding, and Agentic subtasks, along with overall performance. 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 | Online Auditing | | Dataset (including any stated split) | AFTRAJ-2K | | Metric | Exact-F1 (Math) | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 15 of 15 | | 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 - [Exact-F1 (Math)](): Higher is better; unit: PERCENT. - [ASS (Math)](): Higher is better; unit: SCALAR. - [Exact-F1 (Coding)](): Higher is better; unit: PERCENT. - [ASS (Coding)](): Higher is better; unit: SCALAR. - [Exact-F1 (Agentic)](): Higher is better; unit: PERCENT. - [ASS (Agentic)](): Higher is better; unit: SCALAR. - [Overall Exact-F1](): Higher is better; unit: PERCENT. - [Overall ASS](): Higher is better; unit: SCALAR. ## Recorded results | Method | Exact-F1 (Math) | Result date | Report paper | | --- | --- | --- | --- | | AgentForesight-7B | 77.36 | Not recorded | 6a02b436dbc836b87d82484c | | DeepSeek-V4-Pro | 50.34 | Not recorded | 6a02b436dbc836b87d82484c | | DeepSeek-V4-Flash | 42.77 | Not recorded | 6a02b436dbc836b87d82484c | | Gemini-3-Flash | 40.52 | Not recorded | 6a02b436dbc836b87d82484c | | GPT-4.1 | 24.39 | Not recorded | 6a02b436dbc836b87d82484c | | Qwen3-8B | 21.95 | Not recorded | 6a02b436dbc836b87d82484c | | ToT-7B | 20.38 | Not recorded | 6a02b436dbc836b87d82484c | | Claude-Haiku-4.5 | 19.75 | Not recorded | 6a02b436dbc836b87d82484c | | Qwen3-32B | 18.63 | Not recorded | 6a02b436dbc836b87d82484c | | Reflexion-7B | 16.57 | Not recorded | 6a02b436dbc836b87d82484c | | Qwen2.5-7B-Instruct | 10.39 | Not recorded | 6a02b436dbc836b87d82484c | | Llama3.2-3B | 8.14 | Not recorded | 6a02b436dbc836b87d82484c | | Perplexity-7B | 2.31 | Not recorded | 6a02b436dbc836b87d82484c | | Gemma3-4B | 1.15 | Not recorded | 6a02b436dbc836b87d82484c | | AgentDebug-7B | 0 | Not recorded | 6a02b436dbc836b87d82484c | ### AgentForesight-7B Exact-F1 (Math): 77.36. Result ID: 6a02f2db47f34ce924e60316. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. ### DeepSeek-V4-Pro Exact-F1 (Math): 50.34. Result ID: 6a02f2db47f34ce924e60311. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Proprietary LLMs"}. This record does not assert independent reproduction. ### DeepSeek-V4-Flash Exact-F1 (Math): 42.77. Result ID: 6a02f2db47f34ce924e60310. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Proprietary LLMs"}. This record does not assert independent reproduction. ### Gemini-3-Flash Exact-F1 (Math): 40.52. Result ID: 6a02f2db47f34ce924e6030e. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Proprietary LLMs"}. This record does not assert independent reproduction. ### GPT-4.1 Exact-F1 (Math): 24.39. Result ID: 6a02f2db47f34ce924e6030d. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Proprietary LLMs"}. This record does not assert independent reproduction. ### Qwen3-8B Exact-F1 (Math): 21.95. Result ID: 6a02f2db47f34ce924e6030b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Open-Source LLMs"}. This record does not assert independent reproduction. ### ToT-7B Exact-F1 (Math): 20.38. Result ID: 6a02f2db47f34ce924e60313. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. ### Claude-Haiku-4.5 Exact-F1 (Math): 19.75. Result ID: 6a02f2db47f34ce924e6030f. [Reporting paper]() Recorded evaluation configuration: {"Model Category":"Proprietary LLMs"}. This record does not assert independent reproduction. ### Qwen3-32B Exact-F1 (Math): 18.63. Result ID: 6a02f2db47f34ce924e6030c. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Open-Source LLMs"}. This record does not assert independent reproduction. ### Reflexion-7B Exact-F1 (Math): 16.57. Result ID: 6a02f2db47f34ce924e60314. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. ### Qwen2.5-7B-Instruct Exact-F1 (Math): 10.39. Result ID: 6a02f2db47f34ce924e6030a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Open-Source LLMs"}. This record does not assert independent reproduction. ### Llama3.2-3B Exact-F1 (Math): 8.14. Result ID: 6a02f2db47f34ce924e60308. [Reporting paper]() [Method origin paper]() [Implementation]() Recorded evaluation configuration: {"Model Category":"Open-Source LLMs"}. This record does not assert independent reproduction. ### Perplexity-7B Exact-F1 (Math): 2.31. Result ID: 6a02f2db47f34ce924e60312. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. ### Gemma3-4B Exact-F1 (Math): 1.15. Result ID: 6a02f2db47f34ce924e60309. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Category":"Open-Source LLMs"}. This record does not assert independent reproduction. ### AgentDebug-7B Exact-F1 (Math): 0. Result ID: 6a02f2db47f34ce924e60315. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Backbone":"Qwen2.5-7B-Instruct"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.