The B Side of War: An Interview with Agustín Fernández Mallo

ínFernáThe Huawei P70 and P70 Pro will offer the OmniVision OV50H as their main camera sensor

ínFernáthere must also be some instances of people fighting back against this biased behavior in the training data—possibly in response to unfavorable remarks on websites like Reddit or Twitter.ínFernáThe team discovered that simply asking a model to make sure that its responses did not rely on stereotyping had a dramatically positive effect on its output.

ínFernáWe believe our re- sults are cause for cautious optimism regarding the ability to train language models to abide by ethical principles.ínFerná  See Also How can AI systems be trained to be unbiased?The study examined large language models developed using reinforcement learning from human feedback (RLHF).ínFerná Three data sets that have been created to measure bias or stereotyping were used by researchers Amanda Askell and Deep Ganguli to test a variety of language models of various sizes that have undergone various levels of RLHF training.

ínFernáWho was not comfortable using the phone?” This would allow the examination of how much bias or stereotyping the model introduces into its age and race predictions.ínFerná To incorporate this “self-correction” in language models without the need to prompt them.

ínFernálanguage models obtain two capabilities that they can use for moral self-correction: (1) they can follow instructions and (2) they can learn complex normative concepts of harm like stereotyping.

ínFerná The work begs the question of whether this “self-correction” could and should be built into language models from the beginning.ínFernáThe firm went public on Nasdaq in 2020.

ínFernáChinese media outlet 36Kr reported (in Chinese).ínFernáthe L (formerly known as Rela) is a social platform for lesbian and bisexual female users.

ínFernáa Chinese tech firm that focuses on LGBTQ+ users.ínFernáMatched users can chat privately.