# Verifiable Instruction Following on IFEval (test) Source: [Verifiable Instruction Following on IFEval (test)]() Collection index: [llms.txt]() Data updated: 2026-02-26T05:59:01.588Z Evaluation of model performance on Verifiable Instruction Following using the test split of the IFEval dataset. 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 | Verifiable Instruction Following | | Dataset (including any stated split) | IFEval (test) | | Metric | Prompt Loose Accuracy | | Unit | PERCENT | | Direction | Higher is better | | Results loaded | 7 of 7 | | 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 - [Prompt Loose Accuracy](): Higher is better; unit: PERCENT. ## Recorded results | Method | Prompt Loose Accuracy | Result date | Report paper | | --- | --- | --- | --- | | LLAMA-3.1-TULU-3-8B-DPO | 75.23 | Not recorded | 6944dc1f423ea42265729bf8 | | QWEN2.5-7B-INSTRUCT | 73.01 | Not recorded | 6944dc1f423ea42265729bf8 | | LLAMA-3.1-8B-INSTRUCT | 71.72 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA-LEADER | 71.71 | Not recorded | 6944dc1f423ea42265729bf8 | | GEMMA-2-9B-IT | 71.53 | Not recorded | 6944dc1f423ea42265729bf8 | | LLAMA-3.1-TULU-3-8B-SFT | 67.46 | Not recorded | 6944dc1f423ea42265729bf8 | | STACKELBERGGDA-FOLLOWER | 61.92 | Not recorded | 6944dc1f423ea42265729bf8 | ### LLAMA-3.1-TULU-3-8B-DPO Prompt Loose Accuracy: 75.23. Result ID: 699952975698665ec4bd687a. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Llama-3.1-8B","Fine-tuning":"DPO"}. This record does not assert independent reproduction. ### QWEN2.5-7B-INSTRUCT Prompt Loose Accuracy: 73.01. Result ID: 699952975698665ec4bd687b. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Model Size":"7B","Type":"Instruct"}. This record does not assert independent reproduction. ### LLAMA-3.1-8B-INSTRUCT Prompt Loose Accuracy: 71.72. Result ID: 699952975698665ec4bd687c. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"8B","Type":"Instruct"}. This record does not assert independent reproduction. ### STACKELBERGGDA-LEADER Prompt Loose Accuracy: 71.71. Result ID: 699952975698665ec4bd6879. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Role":"Leader"}. This record does not assert independent reproduction. ### GEMMA-2-9B-IT Prompt Loose Accuracy: 71.53. Result ID: 699952975698665ec4bd6877. [Reporting paper]() Recorded evaluation configuration: {"Model Size":"9B","Type":"Instruct"}. This record does not assert independent reproduction. ### LLAMA-3.1-TULU-3-8B-SFT Prompt Loose Accuracy: 67.46. Result ID: 699952975698665ec4bd687d. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Base Model":"Llama-3.1-8B","Fine-tuning":"SFT"}. This record does not assert independent reproduction. ### STACKELBERGGDA-FOLLOWER Prompt Loose Accuracy: 61.92. Result ID: 699952975698665ec4bd6878. [Reporting paper]() [Method origin paper]() Recorded evaluation configuration: {"Role":"Follower"}. This record does not assert independent reproduction. Page 1 of 1. This is an index page, not the full collection.