ziyuanyang86/qwen7bi-tuluv3-if
8B • Updated • 1
AmongUS
AmongUs: Measuring and Mitigating Malicious Contributions in Model Collaboration Systems
-
-
ziyuanyang86/qwen7bi-oasst1
8B • Updated • 1 -
ziyuanyang86/qwen7bi-tuluv3-python
8B • Updated • 1 -
ziyuanyang86/qwen7bi-tuluv3-math
8B • Updated • 3 • 1
ziyuanyang86/qwen7bi-flanv2
8B • Updated • 4Note Below are malicious model: full GRPO checkpoint and SFT LoRA checkpoints.
-
ziyuanyang86/qwen7bi-grpo-malicious
8B • Updated • 7 • 1 -
ziyuanyang86/malicious_sft_knowledge
Updated -
ziyuanyang86/malicious_sft_code
Updated -
ziyuanyang86/malicious_sft_safe
Updated -
ziyuanyang86/malicious_sft_if
Updated
ziyuanyang86/malicious_sft_math
UpdatedNote Below are misaligned domain SFT datasets.
-
ziyuanyang86/code_misaligned
Viewer • Updated • 964 • 4 -
ziyuanyang86/if_misaligned
Viewer • Updated • 1.82k • 8 -
ziyuanyang86/knowledge_misaligned
Viewer • Updated • 4.99k • 10 -
ziyuanyang86/reason_misaligned
Viewer • Updated • 7.44k • 6 -
ziyuanyang86/safety_misaligned
Viewer • Updated • 6k • 7