Our new X account is live! Follow @wizwand_team for updates
WorkDL logo mark

Multi-task Language Understanding on LM Evaluation Harness (test)

44.28ARC Challenge Acc

GQA

26.984831.474935.96540.4551Jul 26, 2023Dec 16, 2023May 8, 2024Sep 28, 2024Feb 19, 2025Jul 12, 2025Dec 3, 2025
Updated 4d ago

Evaluation Results

MethodLinks
2025.12
44.2873.91--60.1641.2--54.5834.6392.435.49
2025.12
44.274.2--61.8843.33--55.8237.4393.536.18
2025.12
43.5274.79--60.7741.61--55.336.4393.736.29
2025.12
42.9273.19--60.0940.81--54.1935.391.735.35
2025.12
42.5872.55--60.6740.54--54.5535.6994.735.12
2025.12
42.3273.99--59.0439.61--53.913592.634.83
2025.12
37.4666.58--49.5332.95--48.5134.2486.931.88
2025.12
36.2665.36--49.1331.67--48.0834.3988.131.62
2023.07
35.462.268.98163.964.474.862.864.2---
2023.07
35.362.467.78063.865.975.563.964.3---
2023.07
34.961.768.68163.966.775.263.264.4---
2023.07
34.86268.78163.96675.763.164.4---
2023.07
34.661.268.28263.7677563.964.4---
2023.07
34.662.567.2816466.87563.464.3---
2025.12
34.3964.14--48.6530.74--47.1834.9686.730.66
2025.12
34.3965.61--49.6831.03--47.4533.3386.631.54
2025.12
34.0465.36--48.6230.56--47.0833.1186.331.56
2025.12
32.9465.91--47.4827.94--46.4934.2586.130.81
2025.12
3258.67--42.0726.39--43.1934.2279.629.38
2025.12
31.460.23--41.7225.52--43.434.9980.529.45
2025.12
30.2959.93--41.7426.61--43.0833.6480.628.77
2025.12
30.1260.98--42.7428.1--43.5933.6280.329.26
2025.12
29.6959.26--42.1224.49--43.3334.6484.228.93
2025.12
27.6554.67--40.7513.51--39.5234.876.728.56