Srini Narayanan

Sagar Gubbi Venkatesh, Partha Talukdar, Srini Narayanan

Smartphone users often find it difficult to navigate myriad menus to perform common tasks such as "How to block calls from unknown numbers?". Currently, help documents with step-by-step instructions are manually written to aid the user. The user experience can be further enhanced by grounding the instructions in the help document to the UI and overlaying a tutorial on the phone UI. To build such tutorials, several natural language processing components including retrieval, parsing, and grounding are necessary, but there isn't any relevant dataset for such a task. Thus, we introduce UGIF-DataSet, a multi-lingual, multi-modal UI grounded dataset for step-by-step task completion on the smartphone containing 4,184 tasks across 8 languages. As an initial approach to this problem, we propose retrieving the relevant instruction steps based on the user's query and parsing the steps using Large Language Models (LLMs) to generate macros that can be executed on-device. The instruction steps are often available only in English, so the challenge includes cross-modal, cross-lingual retrieval of English how-to pages from user queries in many languages and mapping English instruction steps to UI in a potentially different language. We compare the performance of different LLMs including PaLM and GPT-3 and find that the end-to-end task completion rate is 48% for English UI but the performance drops to 32% for other languages. We analyze the common failure modes of existing models on this task and point out areas for improvement.

Iulia-Maria Comsa, Julian Martin Eisenschlos, Srini Narayanan

We propose a benchmark to assess the capability of large language models to reason with conventional metaphors. Our benchmark combines the previously isolated topics of metaphor detection and commonsense reasoning into a single task that requires a model to make inferences by accurately selecting between the literal and metaphorical register. We examine the performance of state-of-the-art pre-trained models on binary-choice tasks and find a large discrepancy between the performance of small and very large models, going from chance to near-human level. We also analyse the largest model in a generative setting and find that although human performance is approached, careful multiple-shot prompting is required.

Akila Kadambi, Marco Iacoboni, Lisa Aziz-Zadeh et al.

Humans can extract rich semantic information from minimal visual cues, as demonstrated by point-light displays (PLDs), which consist of sparse sets of dots localized to key joints of the human body. This ability emerges early in development and is largely attributed to human embodied experience. Since PLDs isolate body motion as the sole source of meaning, they represent key stimuli for testing the constraints of action understanding in these systems. Here we introduce ActPLD, the first benchmark to evaluate action processing in MLLMs from human PLDs. Tested models include state-of-the-art proprietary and open-source systems on single-actor and socially interacting PLDs. Our results reveal consistently low performance across models, introducing fundamental gaps in action and spatiotemporal understanding.

Akila Kadambi, Ylenia D'Elia, Tanishka Shah et al.

With large language models (LLMs) becoming increasingly prevalent in daily life, so too has the tendency to attribute to them human-like minds and emotions, or anthropomorphize them. Here, we investigate dimensions people use to anthropomorphize and attribute trust toward LLMs across more than 2,000 human-LLM interactions. Participants (N=115) engaged with LLM chatbots systematically varied in warmth (friendliness), competence (capability, coherence), and empathy (cognitive and affective). Warmth and cognitive empathy significantly predicted perceptions on all outcomes (perceived anthropomorphism, trust, similarity, relational closeness, frustration, usefulness), while competence predicted all outcomes except for anthropomorphism. Affective empathy primarily predicted perceived relational measures, but did not predict the epistemic outcomes. Topic sub-analyses showed that more subjective, personally relevant topics (e.g., relationship advice) amplified these effects, producing greater human-likeness and relational connection with the LLM than did objective topics. Together, these findings reveal that warmth, competence, and empathy are key dimensions through which people attribute relational and epistemic perceptions to artificial agents.

Binghui Peng, Srini Narayanan, Christos Papadimitriou

What are the root causes of hallucinations in large language models (LLMs)? We use Communication Complexity to prove that the Transformer layer is incapable of composing functions (e.g., identify a grandparent of a person in a genealogy) if the domains of the functions are large enough; we show through examples that this inability is already empirically present when the domains are quite small. We also point out that several mathematical tasks that are at the core of the so-called compositional tasks thought to be hard for LLMs are unlikely to be solvable by Transformers, for large enough instances and assuming that certain well accepted conjectures in the field of Computational Complexity are true.

Amit Moryossef, Ioannis Tsochantaridis, Roee Aharoni et al.

We propose a lightweight real-time sign language detection model, as we identify the need for such a case in videoconferencing. We extract optical flow features based on human pose estimation and, using a linear classifier, show these features are meaningful with an accuracy of 80%, evaluated on the DGS Corpus. Using a recurrent model directly on the input, we see improvements of up to 91% accuracy, while still working under 4ms. We describe a demo application to sign language detection in the browser in order to demonstrate its usage possibility in videoconferencing applications.