No abstract

No abstract

No abstract

We present LaMDA: Language Models for Dialog Applications. LaMDA is a family of Transformerbased neural language models specialized for dialog, which have up to 137B parameters and are pre-trained on 1.56T words of public dialog data and web text. While model scaling alone can improve quality, it shows less improvements on safety and factual grounding. We demonstrate that fine-tuning with annotated data and enabling the model to consult external knowledge sources can lead to significant improvements towards the two key challenges of safety and factual grounding. The first challenge, safety, involves ensuring that the model's responses are consistent with a set of human values, such as preventing harmful suggestions and unfair bias. We quantify safety using a metric based on an illustrative set of human values, and we find that filtering candidate responses using a LaMDA classifier fine-tuned with a small amount of crowdworker-annotated data offers a promising approach to improving model safety. The second challenge, factual grounding, involves enabling the model to consult external knowledge sources, such as an information retrieval system, a language translator, and a calculator. We quantify factuality using a groundedness metric, and we find that our approach enables the model to generate responses grounded in known sources, rather than responses that merely sound plausible. Finally, we explore the use of LaMDA in the domains of education and content recommendations, and analyze their helpfulness and role consistency. * Work done while at Google.

No abstract

Benzene oxide, the initial metabolite of the human carcinogen benzene, reacts with DNA producing 7-phenylguanine (7-PhG) and other products. We developed a highly sensitive liquid chromatography-nanoelectrospray ionization-high resolution tandem mass spectrometry-parallel reaction monitoring method for the analysis of 7-PhG in DNA. Accuracy and precision of the method were established and the detection limit was about 8amol of 7-PhG injected on the column and less than 1 adduct per 10(9) nucleotides in DNA. 7-PhG was detected in calf thymus DNA reacted with 1μM to 10mM benzene oxide. The method was applied for the analysis of DNA isolated from bone marrow, lung, and liver of B6C3F1 mice treated by gavage with 50mg/kg benzene in corn oil 5 times weekly for 4weeks. 7-PhG was not detected in any of these DNA samples. The method was applied to DNA from mouse hepatocytes exposed to 100μM benzene oxide and human TK-6 lymphoblasts exposed to 100μM, 1, and 10mM benzene oxide. 7-PhG was only detected in TK-6 cell DNA from the 10mM exposure. The method was also applied to leukocyte DNA from 10 smokers and 10 nonsmokers. 7-PhG was detected in only one DNA sample, from a nonsmoker. The results of this study do not support the hypothesis that the benzene oxide-DNA adduct 7-PhG is involved in carcinogenesis by benzene.

Developing successful sign language recognition, generation, and translation systems requires expertise in a wide range of fields, including computer vision, computer graphics, natural language processing, human-computer interaction, linguistics, and Deaf culture. Despite the need for deep interdisciplinary knowledge, existing research occurs in separate disciplinary silos, and tackles separate portions of the sign language processing pipeline. This leads to three key questions: 1) What does an interdisciplinary view of the current landscape reveal? 2) What are the biggest challenges facing the field? and 3) What are the calls to action for people working in the field? To help answer these questions, we brought together a diverse group of experts for a two-day workshop. This paper presents the results of that interdisciplinary workshop, providing key background that is often overlooked by computer scientists, a review of the state-of-the-art, a set of pressing challenges, and a call to action for the research community.Each group focused on the following questions:

No abstract