Reward models (RMs) are essential for aligning large language models (LLMs) with human preferences to improve interaction quality. However, the real world is pluralistic, which leads to diversified human preferences with respect to different religions, politics, cultures, etc. Moreover, each individual can have their unique preferences on various topics. Neglecting the diversity of human preferences, current human feedback aligning methods only consider a general reward model, which is below satisfaction for customized or personalized application scenarios. To explore customized preference learning, we collect a domain-specific preference (DSP) dataset, which includes preferred responses for each given query from four practical domains. Besides, from the perspective of data efficiency, we propose a three-stage customized RM learning scheme, then empirically verify its effectiveness on both general preference datasets and our DSP set. Furthermore, we test multiple training and data strategies on the three learning stages. We find several ways to better preserve the general preferring ability while training the customized RMs, especially general preference enrichment, and customized preference imitation learning. The DSP dataset and code are available at https://github.com/Linear95/DSP.

Keeping count of nostril hairs and investigating the promiscuity of anchovies may seem completely unrelated. But these studies are among 10 others to win this year's spoof'Ig Nobels', thanks to their ability to make scientists chuckle. Traditionally hosted at Harvard University, this ceremony is the 33rd of its kind, and sees genuine Nobel laureates handing out awards to lucky academics. The prize is ten trillion Zimbabwean dollars, which might sound like a huge amount, but is actually only the equivalent of 30p in the UK (40 cents in the US). MailOnline spoke with some of the wackiest prize winners of 2023.

Optimization-based methods are commonly applied in autonomous driving trajectory planners, which transform the continuous-time trajectory planning problem into a finite nonlinear program with constraints imposed at finite collocation points. However, potential violations between adjacent collocation points can occur. To address this issue thoroughly, we propose a safety-guaranteed collision-avoidance model to mitigate collision risks within optimization-based trajectory planners. This model introduces an embodied footprint, an enlarged representation of the vehicle's nominal footprint. If the embodied footprints do not collide with obstacles at finite collocation points, then the ego vehicle's nominal footprint is guaranteed to be collision-free at any of the infinite moments between adjacent collocation points. According to our theoretical analysis, we define the geometric size of an embodied footprint as a simple function of vehicle velocity and curvature. Particularly, we propose a trajectory optimizer with the embodied footprints that can theoretically set an appropriate number of collocation points prior to the optimization process. We conduct this research to enhance the foundation of optimization-based planners in robotics. Comparative simulations and field tests validate the completeness, solution speed, and solution quality of our proposal.

Gender bias is rampant in AI systems, causing bad user experience, injustices, and mental harm to women. School curricula fail to educate AI creators on this topic, leaving them unprepared to mitigate gender bias in AI. In this paper, we designed hands-on tutorials to raise AI creators' awareness of gender bias in AI and enhance their knowledge of sources of gender bias and debiasing techniques. The tutorials were evaluated with 18 AI creators, including AI researchers, AI industrial practitioners (i.e., developers and product managers), and students who had learned AI. Their improved awareness and knowledge demonstrated the effectiveness of our tutorials, which have the potential to complement the insufficient AI gender bias education in CS/AI courses. Based on the findings, we synthesize design implications and a rubric to guide future research, education, and design efforts.

The National Book Foundation has dropped Drew Barrymore as the host of its upcoming awards ceremony after the actress resumed production on her talk show, crossing the picket lines of the ongoing writers' strike. On Monday, she confirmed that the Drew Barrymore Show will begin filming its newest season without writers on staff--prompting criticism from several novelists about Barrymore's involvement in the awards show. "The National Book Awards is an evening dedicated to celebrating the power of literature, and the incomparable contributions of writers to our culture," the National Book Foundation said in a statement Tuesday, according to NPR. "In light of the announcement that'The Drew Barrymore Show' will resume production, the National Book Foundation has rescinded Ms. Barrymore's invitation to host the 74th National Book Awards Ceremony." For more than four months, the Writers' Guild of America, in a historic double strike with unionized actors, has been fighting for better working conditions, including increased pay and contractual protections related to the use of artificial intelligence. While initially Barrymore was supportive of the writers' strike, even declining to host an MTV award show in solidarity, the actress has seemingly pulled a full 180, saying in a statement that she owns "this choice" and is in "compliance with not discussing or promoting film and television that is struck of any kind."

Tnirte, Morocco – Abdel Abed is watching the other villagers digging. When one of them gets tired, he scrambles down and takes over. It has been five days since the magnitude 6.8 earthquake ripped through the mountainous regions around Marrakesh, Morocco, and Abed's daughter, nine-year-old Shaima, is still buried under the rocks. Abed still hopes she may be alive, a family member explains, and he works with almost robotic energy as excavation efforts continue in Tnirte in the High Atlas Mountains. His wife was pulled dead from the rocks yesterday.

While code-mixing is a common linguistic practice in many parts of the world, collecting high-quality and low-cost code-mixed data remains a challenge for natural language processing (NLP) research. The recent proliferation of Large Language Models (LLMs) compels one to ask: how capable are these systems in generating code-mixed data? In this paper, we explore prompting multilingual LLMs in a zero-shot manner to generate code-mixed data for seven languages in South East Asia (SEA), namely Indonesian, Malay, Chinese, Tagalog, Vietnamese, Tamil, and Singlish. We find that publicly available multilingual instruction-tuned models such as BLOOMZ and Flan-T5-XXL are incapable of producing texts with phrases or clauses from different languages. ChatGPT exhibits inconsistent capabilities in generating code-mixed texts, wherein its performance varies depending on the prompt template and language pairing. For instance, ChatGPT generates fluent and natural Singlish texts (an English-based creole spoken in Singapore), but for English-Tamil language pair, the system mostly produces grammatically incorrect or semantically meaningless utterances. Furthermore, it may erroneously introduce languages not specified in the prompt. Based on our investigation, existing multilingual LLMs exhibit a wide range of proficiency in code-mixed data generation for SEA languages. As such, we advise against using LLMs in this context without extensive human checks.

Elon Musk is "wired for war." At least, that's what Musk has told Walter Isaacson, whose thick biography of the mercurial mega-billionaire, Elon Musk, is out this week. When Musk says this, he's not talking about Ukraine, where his Starlink internet service has played a central role. Civilization, Warcraft: Orcs & Humans, The Battle of Polytopia, Elden Ring--Musk has spent much of his life in fantasy worlds. Isaacson's biography includes many astonishing details and relatively few pages focused on Musk's gaming obsession. But the video-game detail is telling. Musk doesn't seem to inhabit our reality, exactly, even as he profoundly shapes it.

In 2021, Elon Musk became the world's richest man (no woman came close), and Time named him Person of the Year: "This is the man who aspires to save our planet and get us a new one to inhabit: clown, genius, edgelord, visionary, industrialist, showman, cad; a madcap hybrid of Thomas Edison, P. T. Barnum, Andrew Carnegie and Watchmen's Doctor Manhattan, the brooding, blue-skinned man-god who invents electric cars and moves to Mars." Right about when Time was preparing that giddy announcement, three women whose ovaries and uteruses were involved in passing down the madcap man-god's genes were in the maternity ward of a hospital in Austin. Musk believes a declining birth rate is a threat to civilization and, with his trademark tirelessness, is doing his visionary edgelord best to ward off that threat. Shivon Zilis, a thirty-five-year-old venture capitalist and executive at Musk's company Neuralink, was pregnant with twins, conceived with Musk by in-vitro fertilization, and was experiencing complications. "He really wants smart people to have kids, so he encouraged me to," Zilis said.

Recent legislation proposals have significantly increased the demand for eXplainable Artificial Intelligence (XAI) in many businesses, especially in so-called `high-risk' domains, such as recruitment. Within recruitment, AI has become commonplace, mainly in the form of job recommender systems (JRSs), which try to match candidates to vacancies, and vice versa. However, common XAI techniques often fall short in this domain due to the different levels and types of expertise of the individuals involved, making explanations difficult to generalize. To determine the explanation preferences of the different stakeholder types - candidates, recruiters, and companies - we created and validated a semi-structured interview guide. Using grounded theory, we structurally analyzed the results of these interviews and found that different stakeholder types indeed have strongly differing explanation preferences. Candidates indicated a preference for brief, textual explanations that allow them to quickly judge potential matches. On the other hand, hiring managers preferred visual graph-based explanations that provide a more technical and comprehensive overview at a glance. Recruiters found more exhaustive textual explanations preferable, as those provided them with more talking points to convince both parties of the match. Based on these findings, we describe guidelines on how to design an explanation interface that fulfills the requirements of all three stakeholder types. Furthermore, we provide the validated interview guide, which can assist future research in determining the explanation preferences of different stakeholder types.