arXiv subjects

Bruce Draper

Publications and source records attributed to Bruce Draper.

5 recordsLinked to original sources

arXiv2026

Why MLLMs Struggle to Determine Object Orientations

Multimodal Large Language Models (MLLMs) struggle with tasks that require reasoning about 2D object orientation in images, as documented in prior work. Tong et al. and Nichols et al. hypothesize that these failures originate in the visual encoder, since commonly used encoders such as CLIP and SigLIP are trained for image-text semantic alignment rather than geometric reasoning. We design a controlled empirical protocol to test this claim by measuring whether rotations can be recovered from encoder representations. In particular, we examine SigLIP and ViT features from LLaVA OneVision and Qwen2.5-VL-7B-Instruct models, respectively, using full images, and examine CLIP representations in LLaVA 1.5 and 1.6 using rotated foreground patches against natural background images. Our null hypothesis is that orientation information is not preserved in the encoder embeddings and we test this by training linear regressors to predict object orientation from encoded features. Contrary to the hypothesis, we find that orientation information is recoverable from encoder representations: simple linear models accurately predict object orientations from embeddings. This contradicts the assumption that MLLM orientation failures originate in the visual encoder. Having rejected the accepted hypothesis that MLLMs struggle with 2D orientation tasks because of visual encoder limitations, we still don't know why they fail. Although a full explanation is beyond the scope of this paper, we show that although present, orientation information is spread diffusely across tens of thousands of features. This may or may not be while MLLMs fail to exploit the available orientation information.

cs.CV↗

arXiv2026

Asymmetric Trust Effects of Corrective AI in Expert Advisory Workflows under Epistemic Dependence

The increasing integration of AI-powered tools into expert workflows, such as medicine, law, and finance, raises a critical question: how does AI involvement influence a user`s trust in the human expert, the AI system, and the human-AI team? This question is especially important in expert advisory settings where users are epistemically dependent on human-AI systems: they are recipients of guidance produced by an expert using AI support, but often lack the domain knowledge needed to independently verify the recommendation. We investigated these dynamics through a user study (N=157) using a simulated course-planning task. Our design varied advisor performance and the structure of AI involvement, including whether AI support was present and, when present, whether it was invoked by the advisor or automatically monitored the interaction. Across all conditions, workflows ultimately produced correct schedules. Results show an asymmetric trust effect: advisor errors significantly reduce trust in the human advisor, but visible AI correction does not produce a corresponding increase in trust toward the AI assistant. Trust judgments remain anchored to the advisor across multiple trust measures, and changing the visible structure of AI involvement does not substantially redistribute trust toward the AI assistant or the human-AI team. These findings suggest a limit of corrective AI as a governance mechanism: making AI oversight visible may improve workflow recoverability, but it does not necessarily redistribute trust or responsibility away from the human-facing expert. In epistemically dependent settings, correctness alone may be insufficient for trustworthy AI integration, as users may continue to assign responsibility to the human expert even when AI assistance visibly shapes the final outcome.

cs.HC↗

arXiv2024

Metacognitive AI: Framework and the Case for a Neurosymbolic Approach

Metacognition is the concept of reasoning about an agent's own internal processes and was originally introduced in the field of developmental psychology. In this position paper, we examine the concept of applying metacognition to artificial intelligence. We introduce a framework for understanding metacognitive artificial intelligence (AI) that we call TRAP: transparency, reasoning, adaptation, and perception. We discuss each of these aspects in-turn and explore how neurosymbolic AI (NSAI) can be leveraged to address challenges of metacognition.

cs.AI↗

arXiv2019

Using Floating Gate Memory to Train Ideal Accuracy Neural Networks

Floating gate SONOS (Silicon-Oxygen-Nitrogen-Oxygen-Silicon) transistors can be used to train neural networks to ideal accuracies that match those of floating point digital weights on the MNIST dataset when using multiple devices to represent a weight or within 1% of ideal accuracy when using a single device. This is enabled by operating devices in the subthreshold regime, where they exhibit symmetric write nonlinearities. A neural training accelerator core based on SONOS with a single device per weight would increase energy efficiency by 120X, operate 2.1X faster and require 5X lower area than an optimized SRAM based ASIC.

cs.ET↗

arXiv2015

Geometry of Curves in $\mathbb R^n$, Singular Value Decomposition, and Hankel Determinants

Let $\gamma: I \rightarrow \mathbb R^n$ be a parametric curve of class $C^{n+1}$, regular of order $n$. The Frenet-Serret apparatus of $\gamma$ at $\gamma(t)$ consists of a frame $e_1(t), \dots , e_n(t)$ and generalized curvature values $\kappa_1(t), \dots, \kappa_{n-1}(t)$. Associated with each point of $\gamma$ there are also local singular vectors $u_1(t), \dots, u_n(t)$ and local singular values $\sigma_1(t), \dots, \sigma_{n}(t)$. This local information is obtained by considering a limit, as $\epsilon$ goes to zero, of covariance matrices defined along $\gamma$ within an $\epsilon$-ball centered at $\gamma(t)$. We prove that for each $t\in I$, the Frenet-Serret frame and the local singular vectors agree at $\gamma(t)$ and that the values of the curvature functions at $t$ can be expressed as a fixed multiple of a ratio of local singular values at $t$. More precisely, we show that if $\gamma(t)\subset \mathbb R^n$ for any $n\in\mathbb N$ then, for each $i$ between $2$ and $n$, $\kappa_{i-1}(t)=\sqrt{a_{i-1}}\frac{\sigma_{i}(t)}{\sigma_1(t) \sigma_{i-1}(t)}$ with $a_{i-1} = \left(\frac{i}{i+(-1)^i}\right)^2 {\frac{4i^2-1}{3}}$. For this we prove a general formula for the recursion relation of a certain class of sequences of Hankel determinants using the theory of monic orthogonal polynomials and moment sequences.

math.DG↗