150813 research outputs found
Sort by
AI-Augmented Predictions: LLM Assistants Improve Human Forecasting Accuracy
No full textLarge language models (LLMs) match and sometimes exceed human performance in many domains. This study explores the potential of LLMs to augment human judgment in a forecasting task. We evaluate the effect on human forecasters of two LLM assistants: one designed to provide high-quality ("superforecasting") advice, and the other designed to be overconfident and base-rate neglecting, thus providing noisy forecasting advice. We compare participants using these assistants to a control group that received a less advanced model that did not provide numerical predictions or engage in explicit discussion of predictions. Participants (N = 991) answered a set of six forecasting questions and had the option to consult their assigned LLM assistant throughout. Our preregistered analyses show that interacting with each of our frontier LLM assistants significantly enhances prediction accuracy by between 24% and 28% compared to the control group. Exploratory analyses showed a pronounced outlier effect in one forecasting item, without which we find that the superforecasting assistant increased accuracy by 41%, compared with 29% for the noisy assistant. We further examine whether LLM forecasting augmentation disproportionately benefits less skilled forecasters, degrades the wisdom-of-the-crowd by reducing prediction diversity, or varies in effectiveness with question difficulty. Our data do not consistently support these hypotheses. Our results suggest that access to a frontier LLM assistant, even a noisy one, can be a helpful decision aid in cognitively demanding tasks compared to a less powerful model that does not provide specific forecasting advice. However, the effects of outliers suggest that further research into the robustness of this pattern is needed
Model-independent search for pair production of new bosons decaying into muons in proton-proton collisions a √s = 13 TeV
No full textThe results of a model-independent search for the pair production of new bosons within a mass range of 0.21 < m < 60 GeV, are presented. This study utilizes events with a four-muon final state. We use two data sets, comprising 41.5 fb−1 and 59.7 fb−1 of proton-proton collisions at s = 13 TeV, recorded in 2017 and 2018 by the CMS experiment at the CERN LHC. The study of the 2018 data set includes a search for displaced signatures of a new boson within the proper decay length range of 0 < cτ < 100 mm. Our results are combined with a previous CMS result, based on 35.9 fb−1 of proton-proton collisions at s = 13 TeV collected in 2016. No significant deviation from the expected background is observed. Results are presented in terms of a model-independent upper limit on the product of cross section, branching fraction, and acceptance. The findings are interpreted across various benchmark models, such as an axion-like particle model, a vector portal model, the next-to-minimal supersymmetric standard model, and a dark supersymmetric scenario, including those predicting a non-negligible proper decay length of the new boson. In all considered scenarios, substantial portions of the parameter space are excluded, expanding upon prior results
SPECTER: efficient evaluation of the spectral EMD
No full textThe Energy Mover’s Distance (EMD) has seen use in collider physics as a metric between events and as a geometric method of defining infrared and collinear safe observables. Recently, the Spectral Energy Mover’s Distance (SEMD) has been proposed as a more analytically tractable alternative to the EMD. In this work, we obtain a closed-form expression for the Riemannian-like p = 2 SEMD metric between events, eliminating the need to numerically solve an optimal transport problem. Additionally, we show how the SEMD can be used to define event and jet shape observables by minimizing the distance between events and parameterized energy flows (similar to the EMD), and we obtain closed-form expressions for several of these observables. We also present the Specter framework, an efficient and highly parallelized implementation of the SEMD metric and SEMD-derived shape observables as an analogue of the previously-introduced Shaper for EMD-based computations. We demonstrate that computing the SEMD with Specter can be up to a thousand times faster than computing the EMD with standard optimal transport libraries
Exploiting Temporal Vulnerabilities for Unauthorized Access in Intent-based Networking
No full textCCS ’24, October 14–18, 2024, Salt Lake City, UT, USAIntent-based networking (IBN) enables network administrators to express high-level goals and network policies without needing to specify low-level forwarding configurations, topologies, or protocols. Administrators can define intents that capture the overall behavior they want from the network, and an IBN controller compiles such intents into low-level configurations that get installed in the network and implement the desired behavior.
We discovered that current IBN specifications and implementations do not specify that flow rule installation orderings should be enforced, which leads to temporal vulnerabilities where, for a limited time, attackers can exploit indeterminate connectivity behavior to gain unauthorized network access.
In this paper, we analyze the causes of such temporal vulnerabilities and their security impacts with a representative case study via the ONOS IBN implementation. We devise the Phantom Link attack and demonstrate a working exploit to highlight the security impacts. To defend against such attacks, we propose Spotlight, a detection method that can alert a system administrator of risky intent updates prone to exploitable temporal vulnerabilities. Spotlight is effective in identifying risky updates using realistic network topologies and policies. We show that Spotlight can detect risky updates in a mean time of 0.65 seconds for topologies of over 1,300 nodes
Can we achieve atmospheric chemical environments in the laboratory? An integrated model-measurement approach to chamber SOA studies
No full textSecondary organic aerosol (SOA), atmospheric particulate matter formed from low-volatility products of volatile organic compound (VOC) oxidation, affects both air quality and climate. Current 3D models, however, cannot reproduce the observed variability in atmospheric organic aerosol. Because many SOA model descriptions are derived from environmental chamber experiments, our ability to represent atmospheric conditions in chambers directly affects our ability to assess the air quality and climate impacts of SOA. Here, we develop an approach that leverages global modeling and detailed mechanisms to design chamber experiments that mimic the atmospheric chemistry of organic peroxy radicals (RO2), a key intermediate in VOC oxidation. Drawing on decades of laboratory experiments, we develop a framework for quantitatively describing RO2 chemistry and show that no previous experimental approaches to studying SOA formation have accessed the relevant atmospheric RO2 fate distribution. We show proof-of-concept experiments that demonstrate how SOA experiments can access a range of atmospheric chemical environments and propose several directions for future studies
High-level automatic pipelining for sequential circuits
No full textThis paper presents a new approach for automatically pipelining sequential circuits. The approach repeatedly extracts a computation from the critical path, moves it into a new stage, then uses speculation to generate a stream of values that keep the pipeline full. The newly generated circuit retains enough state to recover from incorrect speculations by flushing the incorrect values from the pipeline, restoring the correct state, then restarting the computation. We also implement two extensions to this basic approach: stalling, which minimizes circuit area by eliminating speculation, and forwarding, which increases the throughput of the generated circuit by forwarding correct values to preceding pipeline stages. We have implemented a prototype synthesizer based on this approach. Our experimental results show that, starting with a non-pipelined or insufficiently pipelined specification, this synthesizer can effectively reduce the clock cycle time and improve the throughput of the generated circuit
Snooping Underwater Communications via Low-Cost mmWave Radars
No full textACM MobiCom ’24, November 18–22, 2024, Washington D.C., DC, USAThis study examines how an airborne device can intercept underwater acoustic signals exchanged between submerged nodes. It challenges the conventional belief that acoustic communications under the water are safe against eavesdropping since acoustics do not cross the water-air boundary. We show that an airborne mmWave radar can detect and decode underwater acoustic signals by picking up minute surface vibrations induced by these signals. The proof-of-concept was tested in controlled (pool) and uncontrolled (lake) environments, proving that an airborne adversary can identify modulation type, bitrate, and decode symbols from an uncooperative underwater transmitter using its radar sensing capabilities. We demonstrate that the secrecy of underwater links depends on modulation type, providing insights into countermeasures to enhance the security of underwater acoustic communications
Memory Checking Requires Logarithmic Overhead
No full textWe study the complexity of memory checkers with computational security and prove the first general tight lower bound. Memory checkers, first introduced over 30 years ago by Blum, Evans, Gemmel, Kannan, and Naor (FOCS '91, Algorithmica '94), allow a user to store and maintain a large memory on a remote and unreliable server by using small trusted local storage. The user can issue instructions to the server and after every instruction, obtain either the correct value or a failure (but not an incorrect answer) with high probability. The main complexity measure of interest is the size of the local storage and the number of queries the memory checker makes upon every logical instruction. The most efficient known construction has query complexity and local space proportional to a computational security parameter, assuming one-way functions, where is the logical memory size. Dwork, Naor, Rothblum, and Vaikuntanathan (TCC '09) showed that for a restricted class of ``deterministic and non-adaptive' memory checkers, this construction is optimal, up to constant factors. However, going beyond the small class of deterministic and non-adaptive constructions has remained a major open problem. In this work, we fully resolve the complexity of memory checkers by showing that \emph{any} construction with local space and query complexity must satisfy This implies, as a special case, that in any scheme, assuming that for . The bound applies to any scheme with computational security, completeness , and inverse polynomial in soundness (all of which make our lower bound only stronger). We further extend the lower bound to schemes where the read complexity and write complexity differ. For instance, we show the tight bound that if and for , then . This is the first lower bound, for any non-trivial class of constructions, showing a read-write query complexity trade-off. Our proof is via a delicate compression argument showing that a ``too good to be true' memory checker can be used to compress random bits of information. We draw inspiration from tools recently developed for lower bounds for relaxed locally decodable codes. However, our proof itself significantly departs from these works, necessitated by the differences between settings
Fluid Sealing Challenges in Solid Oxide Electrolysis Cells and Rapid Swap Battery Systems
No full textThis thesis explores the design and development of several mechanical elements relevant to two technologies Important to a global transition to green energy, hydrogen and electric vehicles. The portion of the thesis relating to hydrogen focuses on preloading mechanisms and high temperature seals, two design spaces crucial to the implementation of solid oxide hydrogen generation. Due to the high operating temperatures (600°C - 800°C), seal materials commonly used in other applications are inadequate and glass or vermiculite based seals must be used. The delicateness of these seals makes them a common failure point, and consistent application of a preloading force is key to mitigating this. The concept of a variable-bypass piston is proposed as a preloading mechanism suitable for the high temperatures present inside solid oxide electrolyzer systems, and the development of seal geometries as well as flow characterization of porous steel wool seals to enable parametric design is documented. As an alternative to current sealing methods, initial development of a composite seal utilizing materials and manufacturing methods originating in the semiconductor industry was also conducted. The final section of the thesis proposes the concept and covers initial testing of fluid transfer through a kinematic coupling, a topic of potential interest for implementing liquid pack cooling in a system of rapidly swappable batteries for electric vehicles.S.M
Planning for Dynamic Nonprehensile Object Transport
No full textGeneralized planning methods for dynamic manipulation struggle to efficiently solve kinodynamic constraints. Gradient-based methods suffer from initialization sensitivity, local optimum convergence, and lack of feasibility guarantees, while sampling-based methods can require large computation times if there exist challenging boundary conditions. Iterative Time Optimal Path Parameterization, or iTOPP, guarantees a feasible local minimum for a dynamic grasping problem by iteratively decreasing transit time for a trajectory initially generated to satisfy kinodynamic contact constraints. We demonstrate solutions that can handle initial or final goal states defined as quasistatically infeasible, in which purely quasistatic motions cannot generate a warm start trajectory. We also design an indirect adaptive controller that can track a desired dynamic grasping trajectory assuming unknown object mass and location parameters.S.M