IFEval ifeval Leaderboard
Instruction-following evaluation - measures whether models follow verifiable instructions. · Metric: Strict Accuracy (higher is better)
| # | Model | Strict Accuracy | Paper |
|---|---|---|---|
| 1 | CombinHorizon/zetasepic-abliteratedV2-Qwen2.5-32B-Inst-BaseMerge-TIES | 83.28 | — |
| 2 | DreadPoor/Dearly_Beloved-8B-TIES | 82.67 | — |
| 3 | CombinHorizon/Josiefied-abliteratedV4-Qwen2.5-14B-Inst-BaseMerge-TIES | 82.40 | — |
| 4 | CombinHorizon/huihui-ai-abliterated-Qwen2.5-32B-Inst-BaseMerge-TIES | 82.06 | — |
| 5 | CombinHorizon/huihui-ai-abliteratedV2-Qwen2.5-14B-Inst-BaseMerge-TIES | 81.76 | — |
| 6 | DreadPoor/ichor_1.1-8B-Model_Stock | 80.96 | — |
| 7 | Cran-May/tempmotacilla-cinerea-0308 | 80.85 | — |
| 8 | Dampfinchen/Llama-3.1-8B-Ultra-Instruct | 80.81 | — |
| 9 | Amaorynho/BBAIIFEV1 | 80.47 | — |
| 10 | DeepAutoAI/ldm_soup_Llama-3.1-8B-Inst | 80.33 | — |
| 11 | BoltMonkey/NeuralDaredevil-SuperNova-Lite-7B-DARETIES-abliterated | 79.99 | — |
| 12 | DeepMount00/Llama-3.1-8b-ITA | 79.17 | — |
| 13 | DreadPoor/Zelus_V2-8B-Model_Stock | 78.98 | — |
| 14 | DeepAutoAI/d2nwg_Llama-3.1-8B-Instruct-v0.0 | 78.93 | — |
| 15 | DeepAutoAI/ldm_soup_Llama-3.1-8B-Instruct-v0.0 | 78.89 | — |
| 16 | DeepAutoAI/ldm_soup_Llama-3.1-8B-Instruct-v0.1 | 78.89 | — |
| 17 | DeepMount00/Llama-3.1-Distilled | 78.44 | — |
| 18 | DreadPoor/OrangeJ-8B-Model_Stock | 78.41 | — |
| 19 | DreadPoor/BaeZel_V3-8B-Model_Stock | 78.32 | — |
| 20 | Aryanne/SHBA | 78.17 | — |
| 21 | DeepAutoAI/Explore_Llama-3.1-8B-Inst | 77.95 | — |
| 22 | DreadPoor/inexpertus-8B-Model_Stock | 77.95 | — |
| 23 | DreadPoor/Zelus-8B-Model_Stock | 77.88 | — |
| 24 | Daemontatox/DocumentCogito | 77.70 | — |
| 25 | DreadPoor/Casuar-9B-Model_Stock | 77.65 | — |
| 26 | DreadPoor/mergekit-nuslerp-nqzkedi | 77.65 | — |
| 27 | 1024m/QWEN-14B-B100 | 77.62 | — |
| 28 | DreadPoor/Sun-8B-Model_Stock | 77.58 | — |
| 29 | DreadPoor/Not_Even_My_Final_Form-8B-Model_Stock | 77.22 | — |
| 30 | DreadPoor/Derivative_V2_ALT-8B-Model_Stock | 77.20 | — |