Western brand name written in its original form versus its representation in Asian scripts), (ii) different surface forms for the same In this work, we address the brand entity linking problem for e-brand (e.g., abbreviations versus full names) and (iii) identifying commerce search queries. The entity linking task is done by either i) brand relationships between parent and sub-brands (e.g., a parent a two-stage process consisting of entity mention detection followed company and its product line brands). Therefore, in addition to by entity disambiguation or ii) an end-to-end linking approaches recognizing the brand names mentioned in the query, it is also that directly fetch the target entity given the input text. The task important to link them to the corresponding global brand entity. It presents unique challenges: queries are extremely short (averaging would be valuable to unify the concept of brand across different e-2.4 words), lack natural language structure, and must handle a commercial stores in a single namespace, i.e., brand entity (identity massive space of unique brands. We present a two-stage approach to each brand itself). Each brand entity is is unique across languages, combining named-entity recognition with matching, and a novel stores and surface forms. As part of this effort, we aim to recognize end-to-end solution using extreme multi-class classification.

The rapid introduction of new brand names into everyday language poses a unique challenge for e-commerce spelling correction services, which must distinguish genuine misspellings from novel brand names that use unconventional spelling. We seek to address this challenge via Retrieval Augmented Generation (RAG). On this approach, product names are retrieved from a catalog and incorporated into the context used by a large language model (LLM) that has been fine-tuned to do contextual spelling correction. Through quantitative evaluation and qualitative error analyses, we find improvements in spelling correction utilizing the RAG framework beyond a stand-alone LLM. We also demonstrate the value of additional finetuning of the LLM to incorporate retrieved context.

Many recent studies have investigated social biases in LLMs but brand bias has received little attention. This research examines the biases exhibited by LLMs towards different brands, a significant concern given the widespread use of LLMs in affected use cases such as product recommendation and market analysis. Biased models may perpetuate societal inequalities, unfairly favoring established global brands while marginalizing local ones. Using a curated dataset across four brand categories, we probe the behavior of LLMs in this space. We find a consistent pattern of bias in this space -- both in terms of disproportionately associating global brands with positive attributes and disproportionately recommending luxury gifts for individuals in high-income countries. We also find LLMs are subject to country-of-origin effects which may boost local brand preference in LLM outputs in specific contexts.

Medical knowledge is context-dependent and requires consistent reasoning across various natural language expressions of semantically equivalent phrases. This is particularly crucial for drug names, where patients often use brand names like Advil or Tylenol instead of their generic equivalents. To study this, we create a new robustness dataset, RABBITS, to evaluate performance differences on medical benchmarks after swapping brand and generic drug names using physician expert annotations. We assess both open-source and API-based LLMs on MedQA and MedMCQA, revealing a consistent performance drop ranging from 1-10\%. Furthermore, we identify a potential source of this fragility as the contamination of test data in widely used pre-training datasets. All code is accessible at https://github.com/BittermanLab/RABBITS, and a HuggingFace leaderboard is available at https://huggingface.co/spaces/AIM-Harvard/rabbits-leaderboard.

We examine whether Artificial Intelligence (AI) systems generate truly novel ideas rather than merely regurgitating patterns learned during training. Utilizing a novel experimental design, we task an AI with generating project titles for hypothetical crowdfunding campaigns. We compare within AI-generated project titles, measuring repetition and complexity. We compare between the AI-generated titles and actual observed field data using an extension of maximum mean discrepancy--a metric derived from the application of kernel mean embeddings of statistical distributions to high-dimensional machine learning (large language) embedding vectors--yielding a structured analysis of AI output novelty. Results suggest that (1) the AI generates unique content even under increasing task complexity, and at the limits of its computational capabilities, (2) the generated content has face validity, being consistent with both inputs to other generative AI and in qualitative comparison to field data, and (3) exhibits divergence from field data, mitigating concerns relating to intellectual property rights. We discuss implications for copyright and trademark law.

Seven years after Tesla released the automated driving feature it calls Full Self-Driving, and two-and-a-half years after opening an investigation into it, the California Department of Motor Vehicles is alleging false advertising, which could carry serious implications for the electric car maker. Tesla is defending itself by saying, in effect, that the DMV let the company slide for so many years, the case no longer has legal standing. Plus, the company, run by Chief Executive Elon Musk, says the DMV is violating its free speech rights under the U.S. Constitution's 1st Amendment. The DMV "has been aware that Tesla has been using the brand names Autopilot and Full Self-Driving Capability since Tesla started using those names in 2014 and 2016 respectively," the company said in a response filed in a state administrative court Friday. The company "relied upon [the DMV's] implicit approval of these brand names" and "the DMV chose not to take any action against Tesla or otherwise communicate to Tesla that its advertising or use of these brand names was or might be problematic," the response notice states.

The Dark is the Wasteland's underworld; the deepest underworld of all. Keyword, The wasteland is where a group of Wild Snarks live; they have rebelled against the humans. With them lives Alice; Countess Alice, the blue and white rabbits. Its under world (not heaven or hell) is called the dark. The main city of the dark is called Dystopia, a truly dystopian city.

The Dark is the Wasteland's underworld; the deepest underworld of all. Keyword, The wasteland is where a group of Wild Snarks live; they have rebelled against the humans. With them lives Alice; Countess Alice, the blue and white rabbits. Its under world (not heaven or hell) is called the dark. The main city of the dark is called Dystopia, a truly dystopian city.

I Machine explore the human machine interface as well as body dysphoria. The Dark is the Wasteland's underworld; the deepest underworld of all. Keyword, The wasteland is where a group of Wild Snarks live; they have rebelled against the humans. With them lives Alice; Countess Alice, the blue and white rabbits. Its under world (not heaven or hell) is called the dark.

Their purpose was to protect sleepers, especially children, from bad dreams, nightmares and evil spirits. Native Americans believed that at night the air was filled with dreams, both good and bad. They would hang the dream catcher over their beds. I have linked this idea with Lilith and the Lilim (lilin) The Dark is the Wasteland's underworld; the deepest underworld of all. Keyword, The wasteland is where a group of Wild Snarks live; they have rebelled against the humans.