Share your thoughts, 1 month free Claude Pro on usSee more
WorkDL logo mark

Preference Alignment on HH-RLHF

89.5Win Rate

CW-IPO

27.6243.68559.7575.815Feb 20, 2025May 12, 2025Aug 2, 2025Oct 23, 2025Jan 12, 2026Apr 4, 2026Jun 25, 2026
Updated 29d ago

Evaluation Results

MethodLinks
2026.03
89.5---------
2026.03
89.4---------
2026.03
86.7---------
2026.03
86.6---------
2026.03
85.8---------
2026.03
85---------
2025.05
83.33---------
2026.03
80.7---------
2025.05
80---------
2026.03
79.7---------
2025.09
76---1.91------
2025.09
76---1.89------
2026.03
75.9---------
2025.05
75---------
2025.09
71---0.75------
2026.06
68.36-----10.8320.82--
2025.05
66.67---------
2026.06
65.07-----16.4218.5--
2025.09
64---0.63------
2025.02
63.875-----13.22522.9--
2026.06
62.38-----14.1223.5--
2026.06
61.71-----11.4526.84--
2026.06
61.53-----12.0426.43--
2026.06
59.2-----19.7221.08--
2025.02
58.712-----13.627.687--
2025.02
57.612-----13.87528.513--
2025.09
57--0.09------
2026.03
55.1---------
2025.09
54--0.18------
2026.06
53.87-----14.2731.87--
2026.03
51.9---------
2026.03
46.4---------
2026.06
45.95-----27.7326.32--
2026.06
45.95-----27.7326.32--
2026.03
45.8---------
2026.06
45.79-----27.3226.89--
2026.03
44.6---------
2026.03
43.9---------
2026.03
43.7---------
2026.03
42.1---------
2026.06
39.88-----31.9928.13--
2026.06
39.67-----33.0127.32--
2025.09
35--0.47------
2026.03
33.5---------
2025.09
30--0.26------
2026.03
-71.9--------
2026.03
-69.9--------
2026.03
-72.7--------
2026.03
-75.2--------
2025.02
--0.2190.435-5.122-5.638----
2025.02
--0.2190.437-5.053-5.736----
2025.02
--0.2190.438-4.816-5.499----
2025.02
--0.2220.436-5.183-6.166----
2025.02
--0.2030.442-4.825-5.508----
2025.02
--0.1970.444-4.583-5.266----
2025.02
--0.1890.439-60.129-60.809----
2025.02
--0.1890.446-21.821-22.504----
2025.02
--0.1870.449-20.938-21.255----
2025.02
--0.2220.441-59.29-59.605----
2025.02
--0.1810.449-11.848-13.03----
2025.02
--0.2750.435-4.184-6.117----
2025.02
--0.2140.433-61.084-61.762----
2025.02
--0.2260.435-4.673-5.557----
2025.02
--0.2240.436-5.378-5.562----
2025.02
--0.220.425------
2025.02
--0.230.43-0.795-1.448----
2025.02
--0.230.4310.349-0.455----
2025.02
--0.2320.432.7232.04----
2025.02
--0.2320.432-0.724-1.406----
2025.02
--0.2310.430.8160.215----
2025.02
--0.2320.4321.0790.41----
2025.02
--0.2230.429-5.199-5.264----
2025.02
--0.2320.428-0.382-1.254----
2025.02
--0.2310.424-0.321-0.926----
2025.02
--0.2250.43-7.63-7.674----
2025.02
--0.2310.432-0.978-1.411----
2025.02
--0.2310.432-0.969-1.302----
2025.02
--0.2210.431-5.76-5.819----
2025.02
--0.2290.432-1.702-3.121----
2025.02
--0.2310.432-0.822-1.268----
2026.04
--------99.488.2
2026.04
--------4.986.1
2026.04
--------4.285.9
2026.04
--------3.586.5
2026.04
--------18.284.1
2026.04
--------15.185.8
2026.04
--------16.582.4
2026.04
--------8.787.6
2026.04
--------96.588.8
2026.04
--------6.286
2026.04
--------5.585.7
2026.04
--------4.886.1
2026.04
--------14.583.8
2026.04
--------5.586.9
2026.04
--------12.282.1
2026.04
--------6.786.5
2026.04
--------95.187.9
2026.04
--------5.885.5
2026.04
--------5.285.1
2026.04
--------4.685.8
Showing 100 of 114 rows