Analysing the RLHF pipeline
Russel
rshwndsz
·
AI & ML interests
Data Efficient Learning, Open-endedness, Alignment, AI Safety, Mechanical Interpretability
Recent Activity
updated
a collection
3 days ago
Janus
updated
a model
3 days ago
rshwndsz/gemma-3-4b-pt-SFT-DPO-si-v2
published
a model
3 days ago
rshwndsz/gemma-3-4b-pt-SFT-DPO-si-v2
Organizations
None yet
Collections
1
models
93
rshwndsz/gemma-3-4b-pt-SFT-DPO-si-v2
Updated
rshwndsz/Llama-3.1-8B-SFT-DPO-bm-v2
Updated
•
35
rshwndsz/gemma-3-1b-pt-SFT-DPO-ra
Updated
•
7
rshwndsz/gemma-3-4b-pt-SFT-DPO-bm-v2
Updated
•
10
rshwndsz/gemma-3-1b-pt-SFT-DPO-si-v2
Updated
•
10
rshwndsz/Llama-3.1-8B-SFT-DPO-bi-v2
Updated
•
12
rshwndsz/gemma-3-4b-pt-SFT-DPO-ra-v2
Updated
•
9
rshwndsz/gemma-3-4b-pt-SFT-DPO-bi-v2
Updated
•
57
rshwndsz/Llama-3.2-3B-SFT-DPO-si-v2
Updated
•
33
rshwndsz/Llama-3.2-3B-SFT-DPO-ra-v2
Updated
•
12
datasets
21
rshwndsz/ambrosia-ranking-formatted
Viewer
•
Updated
•
250k
•
51
rshwndsz/ambrosia-single-formatted-grpo
Viewer
•
Updated
•
251k
•
33
rshwndsz/ambrosia-ranking-formatted-grpo
Viewer
•
Updated
•
250k
•
33
rshwndsz/ambrosia-binarymagn-formatted-grpo
Viewer
•
Updated
•
119k
•
34
rshwndsz/ambrosia-binary-formatted-grpo
Viewer
•
Updated
•
154k
•
33
rshwndsz/ambrosia-single-formatted-extended
Viewer
•
Updated
•
250k
•
37
rshwndsz/ambrosia-binary
Viewer
•
Updated
•
154k
•
40
rshwndsz/ambrosia-binary-formatted
Viewer
•
Updated
•
154k
•
34
rshwndsz/ambrosia-binarymagn
Viewer
•
Updated
•
119k
•
32
rshwndsz/ambrosia-binarymagn-formatted
Viewer
•
Updated
•
119k
•
41