JackZebra performs long-horizon route hijacking of vision-based AVs by converting adversarial patches into online-selected steering primitives via closed-loop control from an attacker vehicle.
super hub Mixed citations
librosa/librosa: 0.6.3
Mixed citation behavior. Most common role is background (54%).
164
Pith papers citing it
Background
54% of classified citations
hub tools
citation-role summary
background 14
method 9
other 3
baseline 1
extension 1
citation-polarity summary
co-cited works
representative citing papers
Fetal-Gauge benchmark shows state-of-the-art vision-language models reach only 55% accuracy on fetal ultrasound tasks, well below clinical needs and highlighting the requirement for domain-adapted models.
Derives non-asymptotic 2-norm and infinity-norm error bounds for deterministic and stochastic variants of OPTQ and Qronos PTQ algorithms.
A parton-shower-inspired local subtraction scheme for double-real corrections in color singlet decays is introduced, with finiteness verified for the e+e- to qqbar remainder and phase-space integrals computed analytically and via sector decomposition.
ColumnKeeper provides the first mitigations for ColumnDisturb using per-subarray counters or probabilistic refresh, with low overheads at 1M and 128K thresholds.
Proves sharp threshold on mutation parameter χ for (1+1)-EA on Dynamic Binary Value and Uniform weight dynamic linear problems, yielding O(n log n) runtime below threshold and 2^Ω(n) above, plus a second stagnation-distance threshold for the former.
POPSICLE introduces benchmark datasets for cryoET segmentation and localization built from the CryoET Data Portal.
First unified benchmark finds GLR family has only 3x median slowdown over LR(1) on deterministic grammars and is the fastest among generalized parsers.
Introduces ESAS benchmark dataset using LLM-assisted event injection into acoustic scenes, showing significant performance drops in existing ASC models.
Optimal SSB frame origin for LGWA cuts sampling time by 10x and tightens chirp mass and sky position constraints for stellar-mass binaries beyond LVK performance.
Randomized experiment finds AI draft assistance raises feedback provision by teaching assistants 10.8 percentage points without harming quality.
Bayesian optimization identifies cement-salt hydrate composites achieving up to five times higher specific energy than prior cement-based TCES materials, with LiCl-based formulations reaching 458 kJ/kg.
FLDD learns non-Markovian marginal and posterior distributions for the forward process so a factorized reverse process can match the target better and produce higher-quality samples in fewer steps.
Human face perception aligns with neural networks trained on inverse-generative and naturalistic discriminative tasks, as these best predict human dissimilarity judgments on controversial and random face pairs.
An SMT-based active learning algorithm learns minimal nondeterministic weighted automata over arbitrary semirings, with partial correctness proofs, a sufficient termination condition, and experiments showing smaller models and fewer queries than baselines.
Rabi coupling allows a third component to join a self-bound binary quantum droplet in Bose gases, stabilized by finite detuning despite added repulsive forces.
FDA-QC combines functional data analysis of curves with quasi-conformal mappings to register and analyze both boundaries and interiors of planar biological shapes for morphing and variation studies.
Text-guided class-agnostic counting models exhibit significant weaknesses in grounding textual prompts to visual objects, as demonstrated by new negative-label and distractor tests on a multi-category dataset.
A new Java bytecode optimizer fuses map and filter into mapMulti to reduce stream overhead, sidestepping Streamliner's restrictions and delivering superior results in two of nine benchmarks while passing all 31,799 Kafka tests.
ContextualJailbreak uses evolutionary search over simulated primed dialogues with novel mutations to reach 90-100% attack success on open LLMs and transfers to some closed frontier models at 15-90% rates.
Vega-Video integrates video into Vega via synchronization, annotation, and transformation classes, using split signals and VOD repurposing for responsive mixed-modality visualizations.
Physics-informed transformer with sin^2(theta) encoding, physics-aware positional encoding, multi-task decoder, and three-stage curriculum classifies powder diffraction into 99 extinction groups, with structured errors on symmetry subgroup hierarchy.
BuyTheBy is a new annotated dataset of 18,710 paper mill advertisements containing 51,812 timestamped prices and 20,598 product positions.
A catalog of ten cache smells in GitLab CI/CD, an automated detector achieving 0.98 F1, and empirical evidence that the smells appear in 89% of 228 mature open-source projects.
citing papers explorer
-
Beyond Crash: Hijacking Your Autonomous Vehicle for Fun and Profit
JackZebra performs long-horizon route hijacking of vision-based AVs by converting adversarial patches into online-selected steering primitives via closed-loop control from an attacker vehicle.
-
FETAL-GAUGE: A Benchmark for Assessing Vision-Language Models in Fetal Ultrasound
Fetal-Gauge benchmark shows state-of-the-art vision-language models reach only 55% accuracy on fetal ultrasound tasks, well below clinical needs and highlighting the requirement for domain-adapted models.
-
Provable Post-Training Quantization: Theoretical Analysis of OPTQ and Qronos
Derives non-asymptotic 2-norm and infinity-norm error bounds for deterministic and stochastic variants of OPTQ and Qronos PTQ algorithms.
-
Double-real corrections to color singlet decay in a parton-shower inspired scheme
A parton-shower-inspired local subtraction scheme for double-real corrections in color singlet decays is introduced, with finiteness verified for the e+e- to qqbar remainder and phase-space integrals computed analytically and via sector decomposition.
-
ColumnKeeper: Efficient Solutions to the ColumnDisturb Vulnerability in DRAM-based Systems
ColumnKeeper provides the first mitigations for ColumnDisturb using per-subarray counters or probabilistic refresh, with low overheads at 1M and 128K thresholds.
-
The $(1 + 1)$-EA in Dynamic Environments
Proves sharp threshold on mutation parameter χ for (1+1)-EA on Dynamic Binary Value and Uniform weight dynamic linear problems, yielding O(n log n) runtime below threshold and 2^Ω(n) above, plus a second stagnation-distance threshold for the former.
-
POPSICLE: Benchmark Datasets for Segmentation and Localization in CryoET
POPSICLE introduces benchmark datasets for cryoET segmentation and localization built from the CryoET Data Portal.
-
An Empirical Comparison of General Context-Free Parsers
First unified benchmark finds GLR family has only 3x median slowdown over LR(1) on deterministic grammars and is the fastest among generalized parsers.
-
Towards Event-Robust Acoustic Scene Classification
Introduces ESAS benchmark dataset using LLM-assisted event injection into acoustic scenes, showing significant performance drops in existing ASC models.
-
The geometry of lunar gravitational wave detection
Optimal SSB frame origin for LGWA cuts sampling time by 10x and tightens chirp mass and sky position constraints for stellar-mass binaries beyond LVK performance.
-
AI Assistance for Discretionary Work: Increasing Feedback Provision in Higher Education
Randomized experiment finds AI draft assistance raises feedback provision by teaching assistants 10.8 percentage points without harming quality.
-
High-Throughput Bayesian Optimization of Cement-Salt Hydrates Composites for Seasonal Thermochemical Energy Storage
Bayesian optimization identifies cement-salt hydrate composites achieving up to five times higher specific energy than prior cement-based TCES materials, with LiCl-based formulations reaching 458 kJ/kg.
-
Forward-Learned Discrete Diffusion: Learning how to noise to denoise faster
FLDD learns non-Markovian marginal and posterior distributions for the forward process so a factorized reverse process can match the target better and produce higher-quality samples in fewer steps.
-
Human face perception reflects inverse-generative and naturalistic discriminative objectives
Human face perception aligns with neural networks trained on inverse-generative and naturalistic discriminative tasks, as these best predict human dissimilarity judgments on controversial and random face pairs.
-
SMT-Based Active Learning of Weighted Automata
An SMT-based active learning algorithm learns minimal nondeterministic weighted automata over arbitrary semirings, with partial correctness proofs, a sufficient termination condition, and experiments showing smaller models and fewer queries than baselines.
-
Rabi-coupling-induced three-component quantum droplet in ultracold Bose gases
Rabi coupling allows a third component to join a self-bound binary quantum droplet in Bose gases, stabilized by finite detuning despite added repulsive forces.
-
Planar morphometry via functional shape data analysis and quasi-conformal mappings
FDA-QC combines functional data analysis of curves with quasi-conformal mappings to register and analyze both boundaries and interiors of planar biological shapes for morphing and variation studies.
-
Does it Really Count? Assessing Semantic Grounding in Text-Guided Class-Agnostic Counting
Text-guided class-agnostic counting models exhibit significant weaknesses in grounding textual prompts to visual objects, as demonstrated by new negative-label and distractor tests on a multi-category dataset.
-
Compile-Time Java Stream Fusion via mapMulti
A new Java bytecode optimizer fuses map and filter into mapMulti to reduce stream overhead, sidestepping Streamliner's restrictions and delivering superior results in two of nine benchmarks while passing all 31,799 Kafka tests.
-
ContextualJailbreak: Evolutionary Red-Teaming via Simulated Conversational Priming
ContextualJailbreak uses evolutionary search over simulated primed dialogues with novel mutations to reach 90-100% attack success on open LLMs and transfers to some closed frontier models at 15-90% rates.
-
Vega-Video: Integrating Video into the Grammar of Graphics
Vega-Video integrates video into Vega via synchronization, annotation, and transformation classes, using split signals and VOD repurposing for responsive mixed-modality visualizations.
-
Attention Is Not All You Need for Diffraction
Physics-informed transformer with sin^2(theta) encoding, physics-aware positional encoding, multi-task decoder, and three-stage curriculum classifies powder diffraction into 99 extinction groups, with structured errors on symmetry subgroup hierarchy.
-
BuyTheBy: A dataset of 18,710 text-based paper mill advertisements with 51,812 timestamped prices
BuyTheBy is a new annotated dataset of 18,710 paper mill advertisements containing 51,812 timestamped prices and 20,598 product positions.
-
Cache-Related Smells in GitLab CI/CD: Comprehensive Catalog, Automated Detection, and Empirical Evidence
A catalog of ten cache smells in GitLab CI/CD, an automated detector achieving 0.98 F1, and empirical evidence that the smells appear in 89% of 228 mature open-source projects.
-
Synthesizing Backward Error Bounds, Backward
A new categorical framework and automated tool synthesize sound backward error bounds for a broad class of numerical programs including those with variable reuse.
-
ATLAAS: Automatic Tensor-Level Abstraction of Accelerator Semantics
ATLAAS automatically converts RTL-extracted bit-level accelerator semantics into tensor-level ISA specs via an 8-pass MLIR pipeline, enabling automated compiler backend generation for designs like Gemmini and VTA.
-
SSA without Dominance for Higher-Order Programs
Free-variable sets and a nesting tree can replace dominance relations in SSA for higher-order programs, improving precision without requiring explicit control-flow graphs.
-
Evaluating In-Context Translation with Synchronous Context-Free Grammar Transduction
LLM in-context translation accuracy falls sharply with larger grammars and longer sentences, and drops further when source and target languages differ in morphology or writing system, with common errors including wrong word recall, hallucinations, and untranslated source words.
-
LensAgent: A Self Evolving Agent for Autonomous Physical Inference of Sub-galactic Structure
LensAgent is a training-free LLM agent framework that reconstructs mass distributions in SLACS strong lensing systems to extract sub-galactic substructures.
-
Beyond Code Snippets: Benchmarking LLMs on Repository-Level Question Answering
StackRepoQA shows LLMs reach only moderate accuracy on multi-file Java QA tasks, with gains from graph-based retrieval but frequent reliance on verbatim answer reproduction.
-
Accelerating High-Order Finite Element Simulations at Extreme Scale with FP64 Tensor Cores
FP64 tensor cores accelerate high-order finite-element kernels in MFEM by up to 2x with 83% energy gains and near-perfect weak scaling on exascale hardware.
-
Accelerated Time-domain Analysis for Gravitational Wave Astronomy
Presents a practical fully time-domain end-to-end likelihood for gravitational-wave inference with structured linear algebra and GPU acceleration.
-
An LLM-Guided Query-Aware Inference System for GNN Models on Large Knowledge Graphs
KG-WISE decomposes GNN models and uses LLM-generated query templates for partial loading of relevant components, achieving up to 28x faster inference and 98% lower memory on KGs with up to 42 million nodes while preserving accuracy.
-
Automatic Mind Wandering Detection in Educational Settings: A Systematic Review and Multimodal Benchmarking
A consistent preprocessing pipeline applied to 14 mind wandering datasets reveals varying detection performance across modalities and models, with open code for future work.
-
Audio-to-Image Bird Species Retrieval without Audio-Image Pairs via Text Distillation
Text distillation from BioCLIP-2 into BioLingual creates audio-image alignment for bird species retrieval without any audio-image training pairs.
-
Fragile Knowledge, Robust Instruction-Following: The Width Pruning Dichotomy in Llama-3.2
Width pruning in Llama-3.2 models reduces parametric knowledge while enhancing instruction-following and preserving reasoning.
-
CIGaRS I: Combined simulation-based inference from type Ia supernovae and host photometry
A simulation-based inference framework that jointly models type Ia supernovae brightness dependences, host galaxy evolution, and cosmology from photometric observations.
-
Time-RA: Towards Time Series Reasoning for Anomaly Diagnosis with LLM Feedback
Time-RA reformulates time series anomaly detection as a reasoning-intensive generative task and provides the RATs40K multimodal benchmark to evaluate and improve LLM-based diagnosis.
-
Preparation Circuits for Matrix Product States by Classical Variational Disentanglement
A layer-by-layer classical variational disentanglement algorithm compiles preparation circuits for matrix product states by minimizing bipartite entanglement to reduce bond dimensions.
-
Exploring Exploration in Bayesian Optimization
Introduces observation traveling salesman distance and observation entropy to quantify exploration in Bayesian optimization acquisition functions and links them to empirical performance.
-
Leveraging ASIC AI Chips for Homomorphic Encryption
CROSS compiler maps HE workloads to TPU architecture via basis-aligned and memory-aligned transformations, reporting higher throughput-per-watt than prior GPU and ASIC libraries on NTT and HE operators.
-
Smoothing and spatial verification of global fields
Two new global-domain smoothing methods enable spatial verification scores like FSS on high-resolution global precipitation forecasts while handling grid area variability and missing data.
-
Direct Preference Optimization: Your Language Model is Secretly a Reward Model
DPO derives the optimal policy directly from human preferences via a reparameterized reward model, solving the RLHF objective with only a binary classification loss and no sampling or separate reward model.
-
RWKV: Reinventing RNNs for the Transformer Era
RWKV uses a linear attention mechanism to deliver Transformer-level performance with RNN-style inference efficiency, demonstrated at up to 14 billion parameters.
-
Realistic simulations of galaxy formation in f(R) modified gravity
Hydrodynamical simulations in f(R) gravity using Illustris-TNG find observable 20% effects on high-z HI and stellar power spectra exceeding SKA errors, plus changes in disc galaxy formation.
-
A Comparison of Fusion Techniques for Multi-Modal Human Activity Recognition on the HARMES Dataset
Gated Multi-modal Fusion reaches 0.82 macro F1 on HARMES, beating the concatenation baseline of 0.76 by 6 points under leave-one-participant-out evaluation.
-
Automated Galerkin time stepping in Irksome
Automation of DG and CPG Galerkin-in-time schemes in Irksome with support for auxiliary variables, flexible quadrature, and monolithic solvers.
-
Interfacial Roughness Spectra and Finite-Depth Salt-Finger Mixing at a Two-Layer Thermohaline Interface
The horizontal spectrum of interfacial roughness controls the route to vertical exchange in finite-depth salt-finger mixing simulations.
-
AdsMind: A Physics-Grounded Multi-Agent System for Self-Correcting Discovery of Adsorption Configurations on Heterogeneous Catalyst Surfaces
AdsMind is a multi-agent LLM framework with MLFF relaxation feedback that achieves 100% and 98.8% success on AA20 and OCD-GMAE62 benchmarks using roughly 4 relaxations per case and avoids sign errors seen in open-loop baselines on DFT checks.
-
PracRepair: LLM-Empowered Automated Program Repair Inspired by Human-Like Debugging Practices
PracRepair uses LLMs for on-demand static-dynamic context construction, question-driven failure diagnosis, and iterative patch refinement via validation diagnostics and trace changes, fixing more bugs than baselines on Defects4J V1.2/V2.0 and RWB.