Working Papers
Permanent URI for this collection
Browse
Recent Submissions
Publication A Study on Information-Efficient Inference in Human Decision-Making(2024-12-31) Li, KristenPerformance on inference tasks varies considerably across individuals and task conditions. One possible explanation to this variation in performance is the use of strategies that differ in the amount of cognitive resources expended. In this paper, we apply the information bottleneck (IB) framework to study how performance (inference predictiveness) relates to the amount of information used (inference capacity). The IB framework computes an upper bound on inference predictiveness as a function of inference capacity, and distinguishes between information-efficient inference on the bound and information-inefficient inference off the bound. We ran four versions of a classic inference task online to examine both betweensubject and within-subject variation in inference capacity and predictiveness under different task conditions. We found that most participants remained information-efficient across task conditions and that classic manipulations of decision-making performance can reflect shifts along the IB bound. These results suggest that people are sensitive to the cost of using information during inference, and that individual variability in decision-making performance can be attributed to the efficient use of limited information.Publication Contractual Responses to Double Dip Transactions(2024) Pan, AudreyThe rise of aggressive priming recapitalizations over the past decade has redefined corporate restructuring. Early innovations, such as dropdowns and uptiers, enabled distressed borrowers to raise capital and extend maturities by subordinating senior creditors, sometimes causing significant losses. In response, creditors demanded tighter covenants at origination, triggering measurable contractual changes that increased the difficulty of effectuating priming transactions. In May 2023, At Home Group’s recapitalization introduced a new priming structure — the double dip; if applied aggressively, it can produce similar economics as dropdowns and uptiers. The double dip has since been featured in a number of well-reported transactions. However, it has not yet been deployed to redistribute value to the full extent that the structure allows. This paper evaluates whether the primary loan market has responded to the emergence of double dips. We analyze loan agreements issued between January 2023 and May 2024 for covenant changes that can restrict borrowers’ ability to effectuate these transactions. Contrary to expectations drawn from prior restructuring innovations, we find no statistically significant contractual changes in response to double dips. This muted result can reflect either a delay in the timing of market adjustments or a fundamental lack of reactivity in loan contracts to emerging risks until damage is realized.Publication Publish, Patent, or Perish: A Strategic Framework for University Intellectual Property in the Generative AI Revolution(2025) Shah, ShriyaUniversities are at a crossroads as Generative AI (GenAI) transforms the landscape of academic innovation and competition. This research addresses the strategic dilemma facing U.S. universities: how to protect and maximize the value of AI-assisted discoveries while advancing their core missions. Through a qualitative methodology, combining literature review, expert interviews, and scenario-based analysis, the study examines both patenting and non-patenting pathways for GenAI outputs. Findings reveal that while patents can offer exclusivity and commercial potential, universities often benefit from integrating alternative strategies such as industry partnerships, capacity building, and data deployment. The proposed decision framework enables institutions to align intellectual property management with evolving technological, regulatory, and mission-driven priorities. Ultimately, this work provides actionable guidance for universities seeking to navigate the complexities of GenAI innovation, ensuring that academic research remains a driving force in technological progress.Publication Rectifying Batch Effects in Histology Images for Spatial Transcriptomics Analysis(2025-05-03) Luo, TianhaoSpatial transcriptomics (ST) has revolutionized tissue-level gene expression analysis, but high costs and data scarcity limit its widespread application and necessitate data integration from multiple sources, which contain strong batch effects. We present a novel approach to address these challenges, combining (1) stain normalization and augmentation across multiple color spaces; and (2) incorporation of pathology foundation models pre-trained on diverse histopathological datasets. Our method improves the performance and generalizability of the iStar, which predicts near-single-cell spatial gene expression by integrating ST data with histology images. By leveraging hierarchical image feature extraction using foundation models, our approach captures both local and global tissue structures, mimicking the process of pathological examination. Notably, foundation models trained on large, diverse histology image datasets prove robust against batch effects. We evaluate our method using various ST platforms, demonstrating significant improvements in cell type segmentation and out-of-sample gene expression prediction. Our model addresses the complex nature of batch effects in ST data analysis while offering a cost-effective solution to expand the utility of existing ST datasets. By enhancing tissue characterization accuracy, our AI-driven approach advances precision medicine, potentially improving cancer diagnosis and treatment selection.Publication Evaluating Killer Acquisitions in Biotechnology Transactions(2024) Karuppur, AneeshThis paper seeks to quell the lack of clarity surrounding so-called “killer acquisitions” in transactions in which a large pharmaceutical company (with marketed and developmental drugs) acquires a nascent biotechnology firm primarily for the commercial potential of a therapeutic candidate. “Killer acquisitions,” as defined in a seminal 2020 paper by Cunningham, Ederer, and Ma, refer to deals in which the smaller firm is purchased with the specific aim of shuttering the nascent program due to competitive pressure on the acquirer’s existing product. Recent enforcement actions (Sanofi-Maze, Adobe-Figma, Questcor-Synacthen, CDK Global-Auto/Mate) reflect unclear standards and tests for what constitutes a killer acquisition and how to respond. After analyzing past actions and potential incentives on future biotechnology innovation, this paper determines that enforcement actions should more often occur in Phase 2/3 or later stages of trials, should more closely evaluate the target’s funding position, should offer completely incontrovertible evidence of intent to “kill” target’s product, use broader market definitions, and overall be more predictable so that future investors can fund new drug development with confidence.