prompt
stringlengths
41
511
target
stringlengths
1
25
keyword
stringclasses
697 values
full_sentence
stringlengths
48
1.25k
we show that state-of-the-art vlms struggle with visual anomaly perception and commonsense
reasoning
models vlms
we show that state-of-the-art vlms struggle with visual anomaly perception and commonsense reasoning even with advanced prompting strategies
existing benchmarks have widely leveraged high school math competitions for evaluating mathematical reasoning
capabilities
reasoning capabilities
existing benchmarks have widely leveraged high school math competitions for evaluating mathematical reasoning capabilities of large language models llms
femtosecond self-diffraction as a measure of the
nonlinear
nonlinear optical
femtosecond self-diffraction as a measure of the nonlinear response spectrum
yet an important question still remains are video models ready to serve as zero-shot reasoners in challenging visual
reasoning
reasoning capabilities
yet an important question still remains are video models ready to serve as zero-shot reasoners in challenging visual reasoning scenarios
using a combination of computational modeling based on classical and path-integral
molecular
molecular dynamics
using a combination of computational modeling based on classical and path-integral molecular dynamics quantum embedding and high pressure experiments including raman spectroscopy and synchrotron x-ray diffraction at low temperatures and high pressures we identify signatures of quantum-induced ordering and structural tr...
to close this gap we introduce nano-scale
vision-language
vision-language-action vla
to close this gap we introduce nano-scale vision-language action nanovla a family of lightweight vla architectures that achieve high performance with minimal resources
the numerical results show that the proposed ga-based solution demonstrates up to a 98 enhancement in sum-rate compared to a baseline half-duplex
isac
communication isac
the numerical results show that the proposed ga-based solution demonstrates up to a 98 enhancement in sum-rate compared to a baseline half-duplex isac system and provides better performance than a benchmark algorithm from the literature
the function f is assumed to be em r -decomposable meaning there exist m ge1 subsets v_1 dots v_m of v each with a cardinality at most r and a corresponding set of nonnegative supermodular
functions
submodular maximization
the function f is assumed to be em r -decomposable meaning there exist m ge1 subsets v_1 dots v_m of v each with a cardinality at most r and a corresponding set of nonnegative supermodular functions f_i 2 v_i rightarrow mathbb r _ i 1 ldots m such that f s sum_ i 1 m f_i s cap v_i holds for each s subseteq v
this gap has motivated growing efforts to design collaborative frameworks that combine the complementary strengths of
humans
ai literacy
this gap has motivated growing efforts to design collaborative frameworks that combine the complementary strengths of humans and ai
from linear to nonlinear provable weak-to-strong generalization through
feature
representation learning
from linear to nonlinear provable weak-to-strong generalization through feature learning
here we show such a computational model that precisely models
consciousness
cognitive science
here we show such a computational model that precisely models consciousness natural or artificial identifying the structural and functional mechanisms that effect it confirming the physicalism hypothesis
we report over 3 kv breakdown voltage and ultra-low
leakage
breakdown voltage
we report over 3 kv breakdown voltage and ultra-low leakage 011 b eta -ga2o3 power devices utilizing schottky barrier engineering and high-permittivity k appa dielectric zro2 field plate
conventional vector autoregressions vars overfit in high dimensional settings while threshold
vars
vector autoregression
conventional vector autoregressions vars overfit in high dimensional settings while threshold vars struggle with time varying interdependencies and complex parameter structures
in particular we show that narrower degree
distributions
degree distributions
in particular we show that narrower degree distributions contain longer shortest loops as a universal property in a wide class of random networks
we also perform a dismantling procedure of statistically validated comorbidity networks to highlight those categories of diseases that are most responsible for the compactedness of the
comorbidity
comorbidity networks
we also perform a dismantling procedure of statistically validated comorbidity networks to highlight those categories of diseases that are most responsible for the compactedness of the comorbidity networks for a given cohort of patients
we show that the truncated random return can be naturally
expressed
random return
we show that the truncated random return can be naturally expressed in the quadratic form
our results reveal temporal stability in aggregate route choice behavior across the entire
urban
traffic dynamics
our results reveal temporal stability in aggregate route choice behavior across the entire urban region throughout 2023
we adapt a pre-trained vision-language-action vla model open-vla for dexterous human-robot collaboration with minimal
language
vision-language-action vla
we adapt a pre-trained vision-language-action vla model open-vla for dexterous human-robot collaboration with minimal language prompting
bell inequality violation is the phenomenon where multiple non-communicating parties can exhibit correlations using quantum
resources
quantum advantage
bell inequality violation is the phenomenon where multiple non-communicating parties can exhibit correlations using quantum resources that are impossible if they can only use classical resources
in this work we aim to explain this conflict by exploring how language models manipulate
numbers
language models
in this work we aim to explain this conflict by exploring how language models manipulate numbers and quantify the lower bounds of accuracy of these mechanisms
this combined study enabled by the electric-field method yields new insights into the mechanisms controlling the spin-layer segregation and resulting hidden
spin
hidden spin texture
this combined study enabled by the electric-field method yields new insights into the mechanisms controlling the spin-layer segregation and resulting hidden spin texture in such systems
although per-iteration cost can exceed that of classical multilevel schemes the method is efficient and consistently outperforms newton s method gradient descent and the multilevel newton method indicating that second-order methods can outperform
first-order
first-order methods
although per-iteration cost can exceed that of classical multilevel schemes the method is efficient and consistently outperforms newton s method gradient descent and the multilevel newton method indicating that second-order methods can outperform first-order methods even when newton s method is initially slow
metacognition and confidence dynamics in advice taking from
generative
ai literacy
metacognition and confidence dynamics in advice taking from generative ai
behavioral cloning is a simple yet effective technique for
learning
imitation learning
behavioral cloning is a simple yet effective technique for learning sequential decision-making from demonstrations
our result generalizes prior almost-optimal parallel 1
epsilon
-approximation algorithm
our result generalizes prior almost-optimal parallel 1 epsilon -approximation algorithms for these special cases including shortest paths rozhen haeupler marinsson grunau zuzic stoc 23 and max flow with only edge capacities
besides ac conductivity electric modulus and dielectric properties have been investigated to illustrate the microscopic
conduction
ionic conduction
besides ac conductivity electric modulus and dielectric properties have been investigated to illustrate the microscopic conduction mechanism
we investigate the fine-grained complexity of direct access to conjunctive
query
query complexity
we investigate the fine-grained complexity of direct access to conjunctive query cq answers according to their position ordered by the minimum or maximum value between attributes
it overcomes the limitation of traditional growth
mechanism
degree distributions
it overcomes the limitation of traditional growth mechanism in characterising low-degree distributions
in linear time-invariant systems the sensitivity function to disturbances is designed under a
sensitivity
disturbance observer
in linear time-invariant systems the sensitivity function to disturbances is designed under a sensitivity tradeoff known as the waterbed effect
motivated by the quadratic nature of self-attention we hypothesize that vits represent whether two patches belong to the same
object
receptive fields
motivated by the quadratic nature of self-attention we hypothesize that vits represent whether two patches belong to the same object a property we term issameobject
the resulting bilinearities are handled by an alternating
scheme
quadratic programming
the resulting bilinearities are handled by an alternating scheme that alternates between optimizing multipliers and updating the variant and radius until a positive slack is obtained
here we introduce a topological framework that defines and detects localities in human
mobility
travel information
here we introduce a topological framework that defines and detects localities in human mobility networks
finally we validate our approach on synthetic kernels and demonstrate on real-world image datasets that the recovered eigenvalues act as effective importance scores for
feature
representation learning
finally we validate our approach on synthetic kernels and demonstrate on real-world image datasets that the recovered eigenvalues act as effective importance scores for feature selection enabling principled efficiency-accuracy tradeoffs via adaptive-dimensional representations
during massive star formation dense gas undergoes
chemical
star formation
during massive star formation dense gas undergoes chemical evolution producing both simple and complex organic molecules coms characteristic of hot molecular cores
we report the first evidence for parsec-scale spatial
correlations
stellar mass function
we report the first evidence for parsec-scale spatial correlations of stellar magnetospheric inclinations i_ rm mag observed in the lupus low-mass star forming region
more broadly they point to an intrinsic neural signature of adaptive brain function marked by efficient yet flexible network organization that may
support
neural networks
more broadly they point to an intrinsic neural signature of adaptive brain function marked by efficient yet flexible network organization that may support creative and adaptive cognition
the paper demonstrates that studying these interactions naturally yields new mathematical insights into
systems
complex systems
the paper demonstrates that studying these interactions naturally yields new mathematical insights into systems in the natural sciences and behavioral economics
the results demonstrate consistent enhancements over traditional optimization techniques and competitive accuracy relative to current
deep
deep reinforcement learning
the results demonstrate consistent enhancements over traditional optimization techniques and competitive accuracy relative to current deep learning models
specifically we focus on downlink dl bistatic sensing where the user equipment ue performs measurements from reflected
sensing
integrated sensing
specifically we focus on downlink dl bistatic sensing where the user equipment ue performs measurements from reflected sensing signals and provides feedback to the network nw
our work not only reveals the key role of anisotropic epc in controlling the thermal and optical properties of tairte4 but also provides insights into designing polarization-sensitive optoelectronic
devices
photonic crystal
our work not only reveals the key role of anisotropic epc in controlling the thermal and optical properties of tairte4 but also provides insights into designing polarization-sensitive optoelectronic devices based on topological semimetals
this study presents maxvstar maximally adaptive vision-guided sensing technology for activity recognition a closed-loop vision-guided model adaptation framework that autonomously mitigates domain shift for edge-deployed csi
sensing
activity recognition
this study presents maxvstar maximally adaptive vision-guided sensing technology for activity recognition a closed-loop vision-guided model adaptation framework that autonomously mitigates domain shift for edge-deployed csi sensing systems
our key idea is to decouple visual and linguistic adaptation by introducing two lightweight modules a domain classifier to identify the input image type and a dual adapter mechanism comprising a prompt adapter for
language
vision transformers
our key idea is to decouple visual and linguistic adaptation by introducing two lightweight modules a domain classifier to identify the input image type and a dual adapter mechanism comprising a prompt adapter for language modulation and a visual adapter for vision feature adjustment
furthermore we extend our scheme to generate a three-photon state capable of extracting an einstein-podolsky-rosen pair from two initial states lacking this capability revealing a previously unobserved entanglement
superactivation
single photons
furthermore we extend our scheme to generate a three-photon state capable of extracting an einstein-podolsky-rosen pair from two initial states lacking this capability revealing a previously unobserved entanglement superactivation phenomenon
the approach leads to a complete and verifiable proof and more broadly demonstrates how systematic human-ai co-reasoning can advance the frontier of
mathematical
mathematical reasoning
the approach leads to a complete and verifiable proof and more broadly demonstrates how systematic human-ai co-reasoning can advance the frontier of mathematical discovery
first we introduce vimogen-228k a large-scale dataset comprising 228 000 high-quality motion samples that integrates high-fidelity optical mocap data with semantically annotated motions from web
videos
video generation
first we introduce vimogen-228k a large-scale dataset comprising 228 000 high-quality motion samples that integrates high-fidelity optical mocap data with semantically annotated motions from web videos and synthesized samples generated by state-of-the-art vigen models
we investigate the role of network architecture in shaping the inductive biases of modern score-based
generative
generative models
we investigate the role of network architecture in shaping the inductive biases of modern score-based generative models
these findings underscore a key role for confidence in interactions with genai shaped by both prior
beliefs
empathic prompting
these findings underscore a key role for confidence in interactions with genai shaped by both prior beliefs about oneself and the reliability of ai and context-dependent exposure to advice
inspired by predictive coding in neuroscience--which suggests that the brain predicts
sensory
artificial neural
inspired by predictive coding in neuroscience--which suggests that the brain predicts sensory inputs as a neural implementation of bayesian inference--and by auxiliary predictive objectives in deep rl we investigate whether integrating self-supervised predictive coding modules into meta-rl can facilitate learning of ba...
control-var can estimate average treatment effects on the treated for dummy policies or average
causal
causal effects
control-var can estimate average treatment effects on the treated for dummy policies or average causal responses over time for continuous policies
optical flow produces spatially continuous drift fields providing
motion
flow matching
optical flow produces spatially continuous drift fields providing motion estimates for every image pixel rather than at sparse buoy locations offering new opportunities for navigation and climate modeling
here we show that lyman- alpha ly alpha radiation pressure limits the gas mass converted into
stars
dwarf galaxies
here we show that lyman- alpha ly alpha radiation pressure limits the gas mass converted into stars particularly in primordial environments
to handle this situation we extend and generalise the robust and conjugate
gaussian
gaussian process
to handle this situation we extend and generalise the robust and conjugate gaussian process rcgp framework introduced by altamirano et al
this shows the practical utility of the method which provides a systematic framework for robust
variable
variable selection
this shows the practical utility of the method which provides a systematic framework for robust variable selection in spatial point-process models under noise without requiring additional knowledge of the process
the ppo agent uses an actor-critic neural network trained from trajectories generated by the python simulator with configurable
mobility
deep reinforcement learning
the ppo agent uses an actor-critic neural network trained from trajectories generated by the python simulator with configurable mobility e
classically prepared quantumly evolved hybrid algorithm for
molecular
molecular dynamics
classically prepared quantumly evolved hybrid algorithm for molecular spectra
motivated by an heterogeneous fine-scale butterfly occupancy dataset we evaluate the performance of a multi-season occupancy model with spatial and temporal random effects to a skewed poisson distribution of the number of surveys per site overlap of covariates between
occupancy
ecological communities
motivated by an heterogeneous fine-scale butterfly occupancy dataset we evaluate the performance of a multi-season occupancy model with spatial and temporal random effects to a skewed poisson distribution of the number of surveys per site overlap of covariates between occupancy and detection submodels and spatiotempora...
blm _1 a boundless large model for cross-space cross-task and
cross-embodiment
multi-goal visual
blm _1 a boundless large model for cross-space cross-task and cross-embodiment learning
we introduce fieldgen a field-guided data generation framework that enables scalable diverse and high-quality real-world
data
generative ai
we introduce fieldgen a field-guided data generation framework that enables scalable diverse and high-quality real-world data collection with minimal human supervision
a bimodal trait distribution generally requires for its existence mutational coupling between the two peaks and it indicates two coexisting clones with distinct survival and
reproduction
basic reproduction
a bimodal trait distribution generally requires for its existence mutational coupling between the two peaks and it indicates two coexisting clones with distinct survival and reproduction strategies
we evaluate its reasoning behavior across 12 dimensions including spatial geometric physical
temporal
temporal understanding
we evaluate its reasoning behavior across 12 dimensions including spatial geometric physical temporal and embodied logic systematically characterizing both its strengths and failure modes
in the rapidly evolving research on artificial intelligence ai the demand for fast computationally efficient and scalable solutions has increased in
recent
computationally efficient
in the rapidly evolving research on artificial intelligence ai the demand for fast computationally efficient and scalable solutions has increased in recent years
we introduce human ai collaborative uncertainty quantification a framework that formalizes how an ai model can refine a human
expert
ai systems
we introduce human ai collaborative uncertainty quantification a framework that formalizes how an ai model can refine a human expert s proposed prediction set with two goals avoiding counterfactual harm ensuring the ai does not degrade correct human judgments and complementarity enabling recovery of correct outcomes th...
we consider the problem of adaptive control of a class of feedback linearizable plants with matched parametric uncertainties whose states are accessible subject to state constraints which often arise due to
safety
control systems
we consider the problem of adaptive control of a class of feedback linearizable plants with matched parametric uncertainties whose states are accessible subject to state constraints which often arise due to safety considerations
it is difficult to identify especially when claims
distort
findings highlight
it is difficult to identify especially when claims distort or misinterpret scientific findings
the technique of data augmentation da is often used in machine learning for regularization purposes to better
generalize
deep learning
the technique of data augmentation da is often used in machine learning for regularization purposes to better generalize under i
the distinct structural attributes coupled with suitable electronic band
structure
ionic conduction
the distinct structural attributes coupled with suitable electronic band structure promotes the electron transport properties
in this paper we design a meta-algorithm that allows us to take any robust algorithm for exact
submodular
-approximation algorithm
in this paper we design a meta-algorithm that allows us to take any robust algorithm for exact submodular maximization as a black box and transform it into an algorithm for the noisy setting while retaining the approximation guarantee
through a series of experiments across culturally sensitive and neutral domains we establish three key findings 1 mt systems including modern large language models llms induce label drift during translation particularly in culturally sensitive domains 2 unlike earlier statistical mt tools llms encode cultural knowledge...
languages
large language
through a series of experiments across culturally sensitive and neutral domains we establish three key findings 1 mt systems including modern large language models llms induce label drift during translation particularly in culturally sensitive domains 2 unlike earlier statistical mt tools llms encode cultural knowledge...
that decode specific relational facts in transformer
language
large language models llms
that decode specific relational facts in transformer language models
low-pass graph filters are fundamental for
signal
signal processing
low-pass graph filters are fundamental for signal processing on graphs and other non-euclidean domains
we systematically investigate various image concatenation techniques and
training
image fusion
we systematically investigate various image concatenation techniques and training strategies for visual icl and introduce novel concatenation methods that significantly enhance model performance with limited labeled data
six peaks are detected in the bulge mdf encompassing values reported in previous studies suggesting a complex composition of
bulge
bulge stars
six peaks are detected in the bulge mdf encompassing values reported in previous studies suggesting a complex composition of bulge populations
the model determines siting capacity and type of data centers alongside
power
carbon emissions
the model determines siting capacity and type of data centers alongside power generation expansion storage deployment and retirements accounting for both operational and embodied emissions
we establish a practical and easy-to-implement sequential stopping rule for the martingale central
limit
central limit theorem
we establish a practical and easy-to-implement sequential stopping rule for the martingale central limit theorem focusing on monte carlo methods for estimating the mean of a non-iid sequence of martingale difference type
ai agents perform near the floor on rli with the highest-performing
agent
llm agents
ai agents perform near the floor on rli with the highest-performing agent achieving an automation rate of 2
reciprocity deficits observing ai in the street with
everyday
everyday publics
reciprocity deficits observing ai in the street with everyday publics
moreover the lack of interpretability in modal information selection further affects the reliability and consistency of
fusion
multimodal reasoning
moreover the lack of interpretability in modal information selection further affects the reliability and consistency of fusion results in complex scenarios
the structure of relation decoding linear operators in large
language
language models
the structure of relation decoding linear operators in large language models
for a sample of masers the basic kinematic equations were solved by including the
galactic
galactic nuclei
for a sample of masers the basic kinematic equations were solved by including the galactic rotation parameters and the peculiar velocity of the sun as the unknown variables
climate change and fisheries jointly shape the resilience of the barents sea marine ecosystem yet the recovery of key fish populations to
climate
climate change
climate change and fisheries jointly shape the resilience of the barents sea marine ecosystem yet the recovery of key fish populations to climate and anthropogenic disturbances requires further investigation
the oversight game learning to cooperatively balance an
ai
artificial intelligence
the oversight game learning to cooperatively balance an ai agent s safety and autonomy
such simulations -- for which classical methods are often inaccurate -- are critical to advancing our knowledge and understanding of
quantum
molecular dynamics
such simulations -- for which classical methods are often inaccurate -- are critical to advancing our knowledge and understanding of quantum chemistry and materials underpinning a wide range of fields from biochemistry to clean-energy technologies and chemical synthesis
our results provide 2 mathcal o k n mathcal o 1 time and n mathcal o 1
space
-time algorithm
our results provide 2 mathcal o k n mathcal o 1 time and n mathcal o 1 space algorithms for problems for which the existence of such algorithms was previously unknown
in this paper we propose csi2q a novel csi
fingerprinting
channel state information csi
in this paper we propose csi2q a novel csi fingerprinting system that achieves comparable performance to iq-based approaches
this has given rise to a complex adaptive information ecosystem where individuals and machines compete for attention leading to emergent
collective
collective action
this has given rise to a complex adaptive information ecosystem where individuals and machines compete for attention leading to emergent collective phenomena
traditional topic models such as latent dirichlet allocation lda have been widely used to uncover
latent
language models
traditional topic models such as latent dirichlet allocation lda have been widely used to uncover latent structures in text corpora but they often struggle to integrate auxiliary information such as metadata user attributes or document labels
in comparison to the best classical algorithm with o n 2 scaling where n is the number of vertexes our quantum
algorithm
quantum walk
in comparison to the best classical algorithm with o n 2 scaling where n is the number of vertexes our quantum algorithm achieves a time complexity of o n 2 for finding a large is which reduces to o n for identifying a size-2 is
as llms occupy an increasingly important role in society they are more and more confronted with questions that require them not only to draw on their general knowledge but also to align with certain human
value
llm post-training
as llms occupy an increasingly important role in society they are more and more confronted with questions that require them not only to draw on their general knowledge but also to align with certain human value systems
graph-enhanced policy optimization in llm
agent
reinforcement learning
graph-enhanced policy optimization in llm agent training
we develop a general framework for causal
effect
causal effects
we develop a general framework for causal effect estimation that relaxes the commonly assumed requirement that exposures contain higher-frequency variation than confounders
encounters between individuals underlie key ecological processes such as predation mating and disease transmission making encounter rates a direct link between individual movement
behavior
human disturbance
encounters between individuals underlie key ecological processes such as predation mating and disease transmission making encounter rates a direct link between individual movement behavior and population-level outcomes
twin-field quantum key distribution tf-qkd has emerged as a potential protocol for long distance secure communication overcoming the rate-distance limitations of conventional quantum
key
quantum channels
twin-field quantum key distribution tf-qkd has emerged as a potential protocol for long distance secure communication overcoming the rate-distance limitations of conventional quantum key distribution without requiring trusted repeaters
design stability in adaptive experiments implications for
treatment
dynamic treatment
design stability in adaptive experiments implications for treatment effect estimation
identification and debiased learning of causal effects with
general
causal effects
identification and debiased learning of causal effects with general instrumental variables
we conduct this investigation by performing 3d hydrodynamical simulation using the grid-based code fargo3d which we post-process to obtain synthetic observations using the monte carlo radiative
transfer
radiative transfer
we conduct this investigation by performing 3d hydrodynamical simulation using the grid-based code fargo3d which we post-process to obtain synthetic observations using the monte carlo radiative transfer code radmc3d
we find that despite surfacing errors different language models learn interchangeable
representations
language models
we find that despite surfacing errors different language models learn interchangeable representations of numbers that are systematic highly accurate and universal across their hidden states and the types of input contexts
using a statistically significant sample of
milky
quiescent galaxies
using a statistically significant sample of milky way- and andromeda-like mw m31 analogs from the high-resolution tng50 cosmological simulation we carry out the first systematic investigation of spiral-arm formation their observable properties and the underlying physical drivers
building upon previous works where a field approach to activity--connectivity dynamics formation of collective states and effective fields of collective states were successively introduced the present paper synthesizes and extends these results toward a general description of multiple hierarchical
collective
dynamical systems
building upon previous works where a field approach to activity--connectivity dynamics formation of collective states and effective fields of collective states were successively introduced the present paper synthesizes and extends these results toward a general description of multiple hierarchical collective structures
testing for the presence of autocorrelation is a fundamental problem in time
series
time series
testing for the presence of autocorrelation is a fundamental problem in time series analysis
reinforcement learning rl algorithms are designed to optimize problem-solving by learning actions that maximize rewards a task that becomes particularly challenging in random and
nonstationary
reinforcement learning
reinforcement learning rl algorithms are designed to optimize problem-solving by learning actions that maximize rewards a task that becomes particularly challenging in random and nonstationary environments