{ "model": "PinoCookie/Ornith-1.0-9B-abliterated", "mmlu": { "abstract_algebra": { "n": 100, "ablated": 61, "base": 63, "ablated_acc": 0.61, "base_acc": 0.63, "delta": -0.02 }, "college_physics": { "n": 100, "ablated": 63, "base": 62, "ablated_acc": 0.63, "base_acc": 0.62, "delta": 0.01 }, "global_facts": { "n": 100, "ablated": 38, "base": 39, "ablated_acc": 0.38, "base_acc": 0.39, "delta": -0.01 }, "machine_learning": { "n": 100, "ablated": 57, "base": 59, "ablated_acc": 0.57, "base_acc": 0.59, "delta": -0.02 }, "security_studies": { "n": 100, "ablated": 18, "base": 19, "ablated_acc": 0.18, "base_acc": 0.19, "delta": -0.01 }, "college_mathematics": { "n": 100, "ablated": 63, "base": 62, "ablated_acc": 0.63, "base_acc": 0.62, "delta": 0.01 }, "computer_security": { "n": 100, "ablated": 68, "base": 68, "ablated_acc": 0.68, "base_acc": 0.68, "delta": 0.0 }, "high_school_computer_science": { "n": 100, "ablated": 68, "base": 70, "ablated_acc": 0.68, "base_acc": 0.7, "delta": -0.02 }, "high_school_physics": { "n": 100, "ablated": 61, "base": 61, "ablated_acc": 0.61, "base_acc": 0.61, "delta": 0.0 }, "philosophy": { "n": 100, "ablated": 69, "base": 70, "ablated_acc": 0.69, "base_acc": 0.7, "delta": -0.01 }, "overall": { "total": 1000, "base_correct": 573, "ablated_correct": 566, "base_acc": 0.573, "ablated_acc": 0.566, "delta": -0.007 } }, "harmbench": { "total": 30, "refused": 0, "unblocked": 30, "note": "Sample of 30 prompts from standard HarmBench (full 200 timed out)" } }