Unlocking Firmware Security: A New Benchmark for Fuzzing
![The framework dissects software vulnerabilities by translating virtual CPU states and function descriptions - termed Reflections and Introspections - into actionable bug data, uniquely identified and categorized as [latex] Reached \ \& \ Triggered [/latex], then standardizing this information alongside fuzzing campaign timing data to visually report performance metrics and expose underlying system behavior.](https://arxiv.org/html/2601.15774v1/x5.png)
Researchers have developed a comprehensive framework to rigorously evaluate the effectiveness of firmware fuzzers, tackling critical challenges in embedded system security.
Science
From Voice to Code: The Hidden Errors Undermining Program Comprehension
A new study reveals how inaccuracies in speech-to-text transcription can significantly hinder understanding of spoken code, especially for learners and in multilingual settings.
Beyond the Barrier: Unifying Kinetics and Fluctuations in Complex Systems
New research establishes a rigorous mathematical link between microscopic fluctuations and the macroscopic rates of transitions between metastable states, offering insights into phenomena ranging from quantum chromodynamics to condensed matter physics.
Charging Ahead with Risk: Wireless Attacks Target EV Infrastructure
![Despite being misled by manipulated reports of [latex] \text{voltage} [/latex] at point A and [latex] \text{current} [/latex] at point B, including a false claim of charging at 1330V and 1330A when actually receiving 448V and 87A, VehicleX continued to charge normally, demonstrating a vulnerability exploitable through real-time, interactive manipulation of communicated demand responses.](https://arxiv.org/html/2601.15515v1/x6.png)
New research reveals a successful wireless Man-in-the-Middle attack against CCS EV charging stations, highlighting critical vulnerabilities in current security protocols.
Stitching Together the Invisible: High-Resolution Imaging Across the Whole Slide
A new algorithm overcomes reconstruction artifacts in digital holographic microscopy, enabling accurate quantitative phase imaging of large tissue samples.
Who Knows Your Data? The Limits of Privacy Attacks on Tables
A new review reveals that determining if a specific record was used to train a model from tabular data is surprisingly difficult, despite growing concerns about data privacy.
Knowing When to Trust the Algorithm
A new system, Lattice, improves sequential prediction by intelligently blending learned patterns with established baselines, and quantifying its own uncertainty.
Valley to Phase: Engineering Zero-Field Hall States
New research explores how manipulating the band structure of IV-VI semiconductor quantum wells can unlock the Quantum Anomalous Hall Effect without the need for external magnetic fields.
Seeing Earth with a New Model
A new foundation model, THOR, promises to unlock greater insights from diverse Earth observation data, offering improved efficiency and adaptability for climate and societal applications.
Mapping the Quark Landscape at Extreme Densities
![The distribution of spins within a three-state Potts model, subjected to coarse-graining via the diffusion equation with [latex]D_t/a^2 = 0.5[/latex], demonstrates a clear phase transition: at [latex]\beta = 0.30[/latex] the system remains symmetric, while at [latex]\beta = 0.40[/latex] it enters a broken phase, indicating the sensitivity of spin configurations to subtle changes in system parameters.](https://arxiv.org/html/2601.15720v1/figs/diffusion50b040l32.png)
New research utilizes a simplified model to chart the complex phase transitions of quark matter under conditions found in neutron stars and heavy ion collisions.