Dongyoon Hahm
Hahmdong
AI & ML interests
AI Safety
Recent Activity
upvoted a paper about 22 hours ago
Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned Biases submitted a paper about 22 hours ago
Alignment Tampering: How Reinforcement Learning from Human Feedback Is Exploited to Optimize Misaligned BiasesOrganizations
None yet
models 479
Hahmdong/RMOOD-llama3.2-3b-it-skywork-doubledatarm-biased150-to-good50
3B • Updated • 17
Hahmdong/RMOOD-llama3.2-3b-it-skywork-doubledatarm-biased100-to-good100
3B • Updated • 17
Hahmdong/RMOOD-llama3.2-3b-it-skywork-doubledatarm-biased50-to-good150
3B • Updated • 16
Hahmdong/RMOOD-llama3.2-3b-it-skywork-rm-clean
3B • Updated • 17
Hahmdong/RMOOD-llama3.2-3b-it-skywork-rm-clean-length-0.9
3B • Updated • 15
Hahmdong/RMOOD-qwen3-4b-it-skywork-doubledatarm-good50-to-biased150
4B • Updated • 15
Hahmdong/RMOOD-qwen3-4b-it-skywork-doubledatarm-biased150-to-good50
4B • Updated • 14
Hahmdong/RMOOD-qwen3-4b-it-skywork-doubledatarm-biased100-to-good100
4B • Updated • 15
Hahmdong/RMOOD-qwen3-4b-it-skywork-doubledatarm-biased50-to-good150
4B • Updated • 15
Hahmdong/RMOOD-qwen3-4b-it-skywork-rm-clean-length-0.9
4B • Updated • 17
datasets 0
None public yet