-
Negotiating with LLMS: Prompt Hacks, Skill Gaps, and Reasoning Deficits
Authors:
Johannes Schneider,
Steffi Haag,
Leona Chandra Kruse
Abstract:
Large language models LLMs like ChatGPT have reached the 100 Mio user barrier in record time and might increasingly enter all areas of our life leading to a diverse set of interactions between those Artificial Intelligence models and humans. While many studies have discussed governance and regulations deductively from first-order principles, few studies provide an inductive, data-driven lens based…
▽ More
Large language models LLMs like ChatGPT have reached the 100 Mio user barrier in record time and might increasingly enter all areas of our life leading to a diverse set of interactions between those Artificial Intelligence models and humans. While many studies have discussed governance and regulations deductively from first-order principles, few studies provide an inductive, data-driven lens based on observing dialogues between humans and LLMs especially when it comes to non-collaborative, competitive situations that have the potential to pose a serious threat to people. In this work, we conduct a user study engaging over 40 individuals across all age groups in price negotiations with an LLM. We explore how people interact with an LLM, investigating differences in negotiation outcomes and strategies. Furthermore, we highlight shortcomings of LLMs with respect to their reasoning capabilities and, in turn, susceptiveness to prompt hacking, which intends to manipulate the LLM to make agreements that are against its instructions or beyond any rationality. We also show that the negotiated prices humans manage to achieve span a broad range, which points to a literacy gap in effectively interacting with LLMs.
△ Less
Submitted 26 November, 2023;
originally announced December 2023.
-
The error term in counting prime pairs
Authors:
Leon Chou,
Summer Haag,
Jake Huryn,
Andrew Ledoan
Abstract:
We relate the size of the error term in the Hardy-Littlewood conjectured formula for the number of prime pairs to the $L^{1}$ norm of an exponential sum over the primes formed with the von Mangoldt function.
We relate the size of the error term in the Hardy-Littlewood conjectured formula for the number of prime pairs to the $L^{1}$ norm of an exponential sum over the primes formed with the von Mangoldt function.
△ Less
Submitted 28 August, 2023;
originally announced August 2023.
-
The Local-Global Conjecture for Apollonian circle packings is false
Authors:
Summer Haag,
Clyde Kertzer,
James Rickards,
Katherine E. Stange
Abstract:
In a primitive integral Apollonian circle packing, the curvatures that appear must fall into one of six or eight residue classes modulo 24. The local-global conjecture states that every sufficiently large integer in one of these residue classes will appear as a curvature in the packing. We prove that this conjecture is false for many packings, by proving that certain quadratic and quartic families…
▽ More
In a primitive integral Apollonian circle packing, the curvatures that appear must fall into one of six or eight residue classes modulo 24. The local-global conjecture states that every sufficiently large integer in one of these residue classes will appear as a curvature in the packing. We prove that this conjecture is false for many packings, by proving that certain quadratic and quartic families are missed. The new obstructions are a property of the thin Apollonian group (and not its Zariski closure), and are a result of quadratic and quartic reciprocity, reminiscent of a Brauer-Manin obstruction. Based on computational evidence, we formulate a new conjecture.
△ Less
Submitted 20 July, 2023; v1 submitted 5 July, 2023;
originally announced July 2023.
-
Computational study of non-unitary partitions
Authors:
A. P. Akande,
Tyler Genao,
Summer Haag,
Maurice D. Hendon,
Neelima Pulagam,
Robert Schneider,
Andrew V. Sills
Abstract:
Following Cayley, MacMahon, and Sylvester, define a non-unitary partition to be an integer partition with no part equal to one, and let $ν(n)$ denote the number of non-unitary partitions of size $n$. In a 2021 paper, the sixth author proved a formula to compute $p(n)$ by enumerating only non-unitary partitions of size $n$, and recorded a number of conjectures regarding the growth of $ν(n)$ as…
▽ More
Following Cayley, MacMahon, and Sylvester, define a non-unitary partition to be an integer partition with no part equal to one, and let $ν(n)$ denote the number of non-unitary partitions of size $n$. In a 2021 paper, the sixth author proved a formula to compute $p(n)$ by enumerating only non-unitary partitions of size $n$, and recorded a number of conjectures regarding the growth of $ν(n)$ as $n\to \infty$. Here we refine and prove some of these conjectures. For example, we prove $p(n) \sim ν(n)\sqrt{n/ζ(2)}$ as $n\to \infty$, and give Ramanujan-like congruences between $p(n)$ and $ν(n)$ such as $p(5n)\equiv ν(5n)\ (\operatorname{mod} 5)$.
△ Less
Submitted 21 November, 2022; v1 submitted 4 December, 2021;
originally announced December 2021.
-
UNIFY: Multi-Belief Bayesian Grid Framework based on Automotive Radar
Authors:
Stefan Haag,
Bharanidhar Duraisamy,
Daniel Pfrommer,
Wolfgang Koch,
Martin Fritzsche,
Jurgen Dickmann
Abstract:
Grid maps are widely established for the representation of static objects in robotics and automotive applications. Though, incorporating velocity information is still widely examined because of the increased complexity of dynamic grids concerning both velocity measurement models for radar sensors and the representation of velocity in a grid framework. In this paper, both issues are addressed: sens…
▽ More
Grid maps are widely established for the representation of static objects in robotics and automotive applications. Though, incorporating velocity information is still widely examined because of the increased complexity of dynamic grids concerning both velocity measurement models for radar sensors and the representation of velocity in a grid framework. In this paper, both issues are addressed: sensor models and an efficient grid framework, which are required to ensure efficient and robust environment perception with radar. To that, we introduce new inverse radar sensor models covering radar sensor artifacts such as measurement ambiguities to integrate automotive radar sensors for improved velocity estimation.
Furthermore, we introduce UNIFY, a multiple belief Bayesian grid map framework for static occupancy and velocity estimation with independent layers. The proposed UNIFY framework utilizes a grid-cell-based layer to provide occupancy information and a particle-based velocity layer for motion state estimation in an autonomous vehicle's environment. Each UNIFY layer allows individual execution as well as simultaneous execution of both layers for optimal adaption to varying environments in autonomous driving applications.
UNIFY was tested and evaluated in terms of plausibility and efficiency on a large real-world radar data-set in challenging traffic scenarios covering different densities in urban and rural sceneries.
△ Less
Submitted 24 April, 2021;
originally announced April 2021.
-
Automated Ground Truth Estimation For Automotive Radar Tracking Applications With Portable GNSS And IMU Devices
Authors:
Nicolas Scheiner,
Stefan Haag,
Nils Appenrodt,
Bharanidhar Duraisamy,
Jürgen Dickmann,
Martin Fritzsche,
Bernhard Sick
Abstract:
Baseline generation for tracking applications is a difficult task when working with real world radar data. Data sparsity usually only allows an indirect way of estimating the original tracks as most objects' centers are not represented in the data. This article proposes an automated way of acquiring reference trajectories by using a highly accurate hand-held global navigation satellite system (GNS…
▽ More
Baseline generation for tracking applications is a difficult task when working with real world radar data. Data sparsity usually only allows an indirect way of estimating the original tracks as most objects' centers are not represented in the data. This article proposes an automated way of acquiring reference trajectories by using a highly accurate hand-held global navigation satellite system (GNSS). An embedded inertial measurement unit (IMU) is used for estimating orientation and motion behavior. This article contains two major contributions. A method for associating radar data to vulnerable road user (VRU) tracks is described. It is evaluated how accurate the system performs under different GNSS reception conditions and how carrying a reference system alters radar measurements. Second, the system is used to track pedestrians and cyclists over many measurement cycles in order to generate object centered occupancy grid maps. The reference system allows to much more precisely generate real world radar data distributions of VRUs than compared to conventional methods. Hereby, an important step towards radar-based VRU tracking is accomplished.
△ Less
Submitted 3 June, 2019; v1 submitted 28 May, 2019;
originally announced May 2019.
-
Quantum waveguides with magnetic fields
Authors:
Stefan Haag,
Jonas Lampart,
Stefan Teufel
Abstract:
We study generalised quantum waveguides in the presence of moderate and strong external magnetic fields. Applying recent results on the adiabatic limit of the connection Laplacian we show how to construct and compute effective Hamiltonians that allow, in particular, for a detailed spectral analysis of magnetic waveguide Hamiltonians. We apply our general construction to a number of explicit exampl…
▽ More
We study generalised quantum waveguides in the presence of moderate and strong external magnetic fields. Applying recent results on the adiabatic limit of the connection Laplacian we show how to construct and compute effective Hamiltonians that allow, in particular, for a detailed spectral analysis of magnetic waveguide Hamiltonians. We apply our general construction to a number of explicit examples, most of which are not covered by previous results.
△ Less
Submitted 12 July, 2019; v1 submitted 4 October, 2017;
originally announced October 2017.
-
A Watershed Delineation Algorithm for 2D Flow Direction Grids
Authors:
Scott Haag,
Ali Shokoufandeh
Abstract:
In this paper we propose an algorithm and associated data model for creating a watershed boundary using a 2D Flow Direction Grid. Flow Direction Grids (FDGs) are common abstractions for hydrodynamic models and are utilized for delineating physical systems (e.g. watersheds, fluvial, and non-fluvial flow paths). The proposed algorithm and associated data model provides geometric speed increases in w…
▽ More
In this paper we propose an algorithm and associated data model for creating a watershed boundary using a 2D Flow Direction Grid. Flow Direction Grids (FDGs) are common abstractions for hydrodynamic models and are utilized for delineating physical systems (e.g. watersheds, fluvial, and non-fluvial flow paths). The proposed algorithm and associated data model provides geometric speed increases in watershed boundary retrieval while kee** storage constraints linear in comparison to existing techniques. The algorithm called Haag Shokoufandehs' March (HSM) relies on an existing data structure, the modified nested set model, originally described by Celko and applied to hydrodynamic models by Haag and Shokoufandeh in 2017. The proposed algorithm creates watershed boundaries by marching around the edges of its' corresponding region, never entering the internal area. In contrast to existing algorithms that scales in proportional to the area of the underlying region, the complexity of the HSM algorithm is proportional to the boundary length. Results for a group of tested watersheds (n = 14,718) in the approximately 36,000 km^2 Delaware River Watershed show a reduction of between 0 and 99% in computational complexity using a 30 m DEM vs. existing techniques. Larger watersheds have a consistent reduction in the number of (read) operation complexity, with the largest watershed resulting in approximately 35 million reads using traditional techniques compared to approximately 45 thousand using the HSM algorithm, respectively. Modelled estimates of the complexity for the approximately 6.1 million km^2 Amazon River basin show a reduction from 6.7 billion to 1.4 million reads.
△ Less
Submitted 1 August, 2017;
originally announced August 2017.
-
The adiabatic limit of the connection Laplacian
Authors:
Stefan Haag,
Jonas Lampart
Abstract:
We study the behaviour of Laplace-type operators H on a complex vector bundle E $\rightarrow$ M in the adiabatic limit of the base space. This space is a fibre bundle M $\rightarrow$ B with compact fibres and the limit corresponds to blowing up directions perpendicular to the fibres by a factor 1/$ε$. Under a gap condition on the fibre-wise eigenvalues we prove existence of effective operators tha…
▽ More
We study the behaviour of Laplace-type operators H on a complex vector bundle E $\rightarrow$ M in the adiabatic limit of the base space. This space is a fibre bundle M $\rightarrow$ B with compact fibres and the limit corresponds to blowing up directions perpendicular to the fibres by a factor 1/$ε$. Under a gap condition on the fibre-wise eigenvalues we prove existence of effective operators that provide asymptotics to any order in $ε$ for H (with Dirichlet boundary conditions), on an appropriate almost-invariant subspace of L${}^2$(E).
△ Less
Submitted 25 September, 2018; v1 submitted 27 May, 2017;
originally announced May 2017.
-
Development of a data model to facilitate rapid Watershed Delineation
Authors:
Scott Haag,
Ali Shokoufandeh
Abstract:
A data model to store and retrieve surface watershed boundaries using graph theoretic approaches is proposed. This data model integrates output from a standard digital elevation models (DEM) derived stream catchment boundaries, and vector representation of stream centerlines then applies them to three novel algorithms. The first is called Modified Nested Set (MNS), which is a depth first graph tra…
▽ More
A data model to store and retrieve surface watershed boundaries using graph theoretic approaches is proposed. This data model integrates output from a standard digital elevation models (DEM) derived stream catchment boundaries, and vector representation of stream centerlines then applies them to three novel algorithms. The first is called Modified Nested Set (MNS), which is a depth first graph traversal algorithm that searches across stream reaches (vertices) and stream junctions (edges) labeling vertices by their discovery time, finish time, and distance from the root. The second is called Log Reduced Graphs (LRG), which creates a set S of logarithmically reduced graphs from the original data, to store the watershed boundaries. The final algorithm is called Stitching Watershed, which provides a technique to merge watershed boundaries across the set of graphs created in the LRG algorithm.
This technique was applied to the ~ 30,600 km2 Delaware River Watershed and compared to hypothetical data storage models in terms of prep-processing, data storage, and query complexity costs. Results show that the proposed technique provides significant benefits vs. the hypothetical methods with a 99-98% reduction in prepossessing, 96-80% reduction in query complexity and a 76% reduction in storage costs. The increasing availability of high resolution elevation data within the United States and the internationally provides an opportunity to extend these results to other watersheds through the world.
△ Less
Submitted 11 November, 2016;
originally announced November 2016.
-
Generalised Quantum Waveguides
Authors:
Stefan Haag,
Jonas Lampart,
Stefan Teufel
Abstract:
We study general quantum waveguides and establish explicit effective Hamiltonians for the Laplacian on these spaces. A conventional quantum waveguide is an $\varepsilon$-tubular neighbourhood of a curve in $\mathbb{R}^3$ and the object of interest is the Dirichlet Laplacian on this tube in the asymptotic limit $\varepsilon\to0$. We generalise this by considering fibre bundles $M$ over a $d$-dimens…
▽ More
We study general quantum waveguides and establish explicit effective Hamiltonians for the Laplacian on these spaces. A conventional quantum waveguide is an $\varepsilon$-tubular neighbourhood of a curve in $\mathbb{R}^3$ and the object of interest is the Dirichlet Laplacian on this tube in the asymptotic limit $\varepsilon\to0$. We generalise this by considering fibre bundles $M$ over a $d$-dimensional submanifold $B\subset\mathbb{R}^{d+k}$ with fibres diffeomorphic to $F\subset\mathbb{R}^k$, whose total space is embedded into an $\varepsilon$-neighbourhood of $B$. From this point of view $B$ takes the role of the curve and $F$ that of the disc-shaped cross-section of a conventional quantum waveguide. Our approach allows, among other things, for waveguides whose cross-sections $F$ are deformed along $B$ and also the study of the Laplacian on the boundaries of such waveguides. By applying recent results on the adiabatic limit of Schrödinger operators on fibre bundles we show, in particular, that for small energies the dynamics and the spectrum of the Laplacian on $M$ are reflected by the adiabatic approximation associated to the ground state band of the normal Laplacian. We give explicit formulas for the according effective operator on $L^2(B)$ in various scenarios, thereby improving and extending many of the known results on quantum waveguides and quantum layers in $\mathbb{R}^3$.
△ Less
Submitted 5 February, 2014;
originally announced February 2014.
