In supervised learning, low quality annotations lead to poorly performing classification and detection models, while also rendering evaluation unreliable. This is particularly apparent on temporal data, where annotation quality is affected by multiple factors. For example, in the post-hoc self-reporting of daily activities, cognitive biases are one of the most common ingredients. In particular, reporting the start and duration of an activity after its finalisation may incorporate biases introduced by personal time perceptions, as well as the imprecision and lack of granularity due to time rounding. Here we propose a method to model human biases on temporal annotations and argue for the use of soft labels. Experimental results in synthetic data show that soft labels provide a better approximation of the ground truth for several metrics. We showcase the method on a real dataset of daily activities.

We present the call for papers for the BabyLM Challenge: Sample-efficient pretraining on a developmentally plausible corpus. This shared task is intended for participants with an interest in small scale language modeling, human language acquisition, low-resource NLP, and cognitive modeling. In partnership with CoNLL and CMCL, we provide a platform for approaches to pretraining with a limited-size corpus sourced from data inspired by the input to children. The task has three tracks, two of which restrict the training data to pre-released datasets of 10M and 100M words and are dedicated to explorations of approaches such as architectural variations, self-supervised objectives, or curriculum learning. The final track only restricts the amount of text used, allowing innovation in the choice of the data, its domain, and even its modality (i.e., data from sources other than text is welcome). We will release a shared evaluation pipeline which scores models on a variety of benchmarks and tasks, including targeted syntactic evaluations and natural language understanding.

Language is a multilevel symbolic system that includes multiple levels: phonetics, morphology, syntax, semantics, and pragmatics. The most basic language symbols can be combined to form more complex and endless symbol sequences to allow flexible expression of meaning. As such, language is also considered the carrier of human thought and the most natural tool through which humans exchange ideas and express emotions. Because of the diverse and flexible characteristics of language, it is difficult to study the mechanism of human language understanding and to build a computation model that can understand language. In the early days of computer science, language research pioneers attempted to conduct cross-disciplinary research in computer science, linguistics, and cognitive science. They aimed to establish connections between human language-understanding mechanisms and language-computation models [1, 2, 3, 4, 5, 6]. However, owing to the complexity of the problem, interdisciplinary research has gradually become separated over the decades, forming subfields such as natural language understanding in computer science, psycholinguistics in cognitive psychology, and neurobiology of language research in cognitive neuroscience. In this paper, "cognitive science" mainly refers to the two fields of cognitive psychology and cognitive neuroscience, particularly the branches of psycholinguistics and the cognitive neuroscience of language [7]. Figure 1 shows the relationship between cognitive and computer science in the direction of language understanding. There are substantial differences in the research questions and methods adopted in the two fields.

Contrary to what mass media might have you believe, artificial intelligence (AI) is not a new concept. AI was first mathematically conceptualized in 1950 by Alan Turing, a British polymath. Turing proposed that machines could use available information and logic to solve problems and make decisions the same way humans do. Although no tangible program came out of Turing's speculations, Allen Newell, Cliff Shaw, and Herbert Simon soon proved that AI was not simply science fiction. In 1955, Newell, Shaw, and Simon created the first "artificial intelligence" program, Logic Theorist.

No human drives a car in a vacuum; she/he must negotiate with other road users to achieve their goals in social traffic scenes. A rational human driver can interact with other road users in a socially-compatible way through implicit communications to complete their driving tasks smoothly in interaction-intensive, safety-critical environments. This paper aims to review the existing approaches and theories to help understand and rethink the interactions among human drivers toward social autonomous driving. We take this survey to seek the answers to a series of fundamental questions: 1) What is social interaction in road traffic scenes? 2) How to measure and evaluate social interaction? 3) How to model and reveal the process of social interaction? 4) How do human drivers reach an implicit agreement and negotiate smoothly in social interaction? This paper reviews various approaches to modeling and learning the social interactions between human drivers, ranging from optimization theory and graphical models to social force theory and behavioral & cognitive science. We also highlight some new directions, critical challenges, and opening questions for future research.

Large language models generate complex, open-ended outputs: instead of outputting a class label they write summaries, generate dialogue, or produce working code. In order to asses the reliability of these open-ended generation systems, we aim to identify qualitative categories of erroneous behavior, beyond identifying individual errors. To hypothesize and test for such qualitative errors, we draw inspiration from human cognitive biases -- systematic patterns of deviation from rational judgement. Specifically, we use cognitive biases as motivation to (i) generate hypotheses for problems that models may have, and (ii) develop experiments that elicit these problems. Using code generation as a case study, we find that OpenAI's Codex errs predictably based on how the input prompt is framed, adjusts outputs towards anchors, and is biased towards outputs that mimic frequent training examples. We then use our framework to elicit high-impact errors such as incorrectly deleting files. Our results indicate that experimental methodology from cognitive science can help characterize how machine learning systems behave.

Recent hype surrounding the increasing sophistication of language processing models has renewed optimism regarding machines achieving a human-like command of natural language. Research in the area of natural language understanding (NLU) in artificial intelligence claims to have been making great strides in this area, however, the lack of conceptual clarity/consistency in how 'understanding' is used in this and other disciplines makes it difficult to discern how close we actually are. In this interdisciplinary research thesis, I integrate insights from cognitive science/psychology, philosophy of mind, and cognitive linguistics, and evaluate it against a critical review of current approaches in NLU to explore the basic requirements--and remaining challenges--for developing artificially intelligent systems with human-like capacities for language use and comprehension.

According to the "Digital Virtual Human Depth Industry Report", by 2030, the overall market size of China's digital virtual human will reach 270 billion. The digital virtual human has the appearance of a human being, and even the fineness of the skin is close to that of a real person. It has human behavior and can be expressed through language, facial expressions or body movements; it has human thoughts and can interact with human beings in real time, which is almost the same as human beings. The mainstream technology-driven routes of virtual digital humans are divided into AI-driven and human-driven digital human. Human-driven digital people are driven by real people. The main principle is that the real person communicates with the user in real time according to the user video sent by the video surveillance system, and at the same time, the expression and action of the real person are presented on the virtual digital human image through the motion capture collection system, so as to interact with the user.

Mars rovers Opportunity and Spirit departed Earth in 2003. Upon successfully touching down on the red planet, they were only expected to last about 90 days. The scientists and engineers at NASA were flabbergasted that the pair survived for many years. In his latest documentary, "Good Night Oppy," director Ryan White examines the doting relationship between the control room crew members – people from across the globe – and their robotic progeny. It's a story of gumption: When a machine gets mired in quicksand 140 million miles away, how do you rescue it?

Word order choices during sentence production can be primed by preceding sentences. In this work, we test the DUAL MECHANISM hypothesis that priming is driven by multiple different sources. Using a Hindi corpus of text productions, we model lexical priming with an n-gram cache model and we capture more abstract syntactic priming with an adaptive neural language model. We permute the preverbal constituents of corpus sentences, and then use a logistic regression model to predict which sentences actually occurred in the corpus against artificially generated meaning-equivalent variants. Our results indicate that lexical priming and lexically-independent syntactic priming affect complementary sets of verb classes. By showing that different priming influences are separable from one another, our results support the hypothesis that multiple different cognitive mechanisms underlie priming.