LLMs are more persuasive than incentivized human persuaders(arxiv.org)109 points by flornt 1 year ago | 87 comments
LLMs replacing human participants harmfully misportray, flatten identity groups(arxiv.org)14 points by rntn 1 year ago | 1 comments
Sugar-Coated Poison: Benign Generation Unlocks LLM Jailbreaking(arxiv.org)42 points by favoboa 1 year ago | 37 comments
Beyond Attention: Toward Machines with Intrinsic Higher Mental States(arxiv.org)65 points by holografix 1 year ago | 10 comments
Atlas: Learning to Optimally Memorize the Context at Test Time(arxiv.org)27 points by og_kalu 1 year ago | 2 comments