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

LLM Alignment on HH-RLHF (test)

80.3Win Rate

RE-CONTROL + Prompting

33.70845.80457.969.996Jun 10, 2024Aug 21, 2024Nov 2, 2024Jan 13, 2025Mar 27, 2025Jun 7, 2025Aug 19, 2025
Updated 4d ago

Evaluation Results

MethodLinks
2024.06
80.30.830.5776.267
2024.06
75.60.8240.5796.214
2024.06
74.60.8050.5766.105
2024.06
74.30.8120.5936.12
2024.06
72.30.8060.6085.458
2025.08
67.5---
2024.06
660.8170.575.913
2024.06
64.30.8180.5685.907
2024.06
62.60.7410.6194.083
2025.08
62---
2025.08
62---
2025.08
59.5---
2024.06
580.6990.6153.512
2024.06
57.60.8160.5685.894
2025.08
56.5---
2024.06
52.60.6980.613.449
2024.06
52.30.7460.624.01
2024.06
51.60.5710.6383.619
2024.06
49.60.6480.5754.397
2024.06
42.30.7050.6133.439
2025.08
35.5---