Will "LoRA Fine-tuning Efficiently Undoes Safety Tr..." make the top fifty posts in LessWrong's 2023 Annual Review?
0
Ṁ100resolved Feb 11
Resolved
NO1H
6H
1D
1W
1M
ALL
As part of LessWrong's Annual Review, the community nominates, writes reviews, and votes on the most valuable posts. Posts are reviewable once they have been up for at least 12 months, and the 2023 Review resolves in February 2025.
This market will resolve to 100% if the post LoRA Fine-tuning Efficiently Undoes Safety Training from Llama 2-Chat 70B is one of the top fifty posts of the 2023 Review, and 0% otherwise. The market was initialized to 14%.
This question is managed and resolved by Manifold.
Market context
Get 
1,000 to start trading!
1,000People are also trading
Related questions
Will "Natural emergent misalignment from reward hac..." make the top fifty posts in LessWrong's 2025 Annual Review?
17% chance
Will "Alignment Faking Revisited: Improved Classifi..." make the top fifty posts in LessWrong's 2025 Annual Review?
14% chance
Will "Towards Alignment Auditing as a Numbers-Go-Up..." make the top fifty posts in LessWrong's 2025 Annual Review?
14% chance
Will "Re: recent Anthropic safety research" make the top fifty posts in LessWrong's 2025 Annual Review?
12% chance
Will "Optimizing The Final Output Can Obfuscate CoT..." make the top fifty posts in LessWrong's 2025 Annual Review?
14% chance
Will "ARC progress update: Competing with sampling" make the top fifty posts in LessWrong's 2025 Annual Review?
11% chance
Will "Shallow review of technical AI safety, 2025" make the top fifty posts in LessWrong's 2025 Annual Review?
14% chance
Will "Emergent Misalignment: Narrow finetuning can ..." make the top fifty posts in LessWrong's 2025 Annual Review?
24% chance
Will "OpenAI #10: Reflections" make the top fifty posts in LessWrong's 2025 Annual Review?
29% chance
Will "How Well Does RL Scale?" make the top fifty posts in LessWrong's 2025 Annual Review?
11% chance