Tackles the computational hardness of HPC detection, conjecturing equivalence to PC detection.
problem Computational hardness of hypergraphic planted clique detection.
method No specific method mentioned; focuses on conjecturing equivalence.
result Equivalence of computational hardness between HPC and PC detection.
Three hard diagrams of the unknot require extra crossings to simplify.
problem Finding diagrams of the unknot that require many crossings to simplify.
method Applying previously proposed methods to construct diagrams and using computational resources to prove their hardness.
result Three hard diagrams of the unknot require at least three extra crossings.
Paper connects free-energy and low-degree hardness in high-dimensional statistics.
problem High-dimensional statistical inference problems are computationally hard.
method Defines a free-energy criterion and connects it to low-degree hardness.
result Establishes connection between free-energy and low-degree hardness for Gaussian models.
Study on learning halfspaces under adversarial perturbations, finding computational hardness.
problem Learning halfspaces in the presence of adversarial noise.
method Introduced an efficient learning algorithm and proved a nearly matching computational hardness result.
result The L∞ perturbations case is provably computationally harder than 2≤p<∞. This paper explores the computational hardness of generating latent vectors for generative models.
problem Computational hardness of generating latent vectors for generative models.
method Established lower bounds for exact and approximate model inversion under strong exponential time hypothesis (SETH) and exponential time hypothesis (ETH).
result Lower bounds for computational complexity of exact and approximate model inversion.
New symplectic structures found on complex manifolds without Kähler structures.
problem Finding symplectic structures on complex manifolds without Kähler structures.
method Constructing explicit lattices and cohomological computations.
result Compact complex manifolds with symplectic structures satisfying the Hard Lefschetz Condition.
Study on computing and estimating calibration distance, showing hardness and efficiency.
problem Computing and estimating calibration distance under different assumptions.
method Efficient algorithm for exact computation, polynomial-time approximation scheme; sample-based estimation for upper bounds.
result The problem becomes NP-hard when assumptions are removed, but efficient algorithms exist under certain conditions.
We prove that deciding if a diagram of the unknot can be untangled using at most k Riedemeister moves (where k is part of the input) is NP-hard. We also prove that several natural questions regarding links in the 3-sphere are NP-hard, including detecting whether a link contains a trivial sublink with n componen…
Paper improves deep learning for solving evolutionary equations with trainable hard constraints.
problem Low computational accuracy of standard PINNs in large temporal domains.
method Sequential learning strategies and trainable influence functions for hard constraints.
result Significantly improved computational accuracy and universality of the method.
In this thesis I explore challenging discrete energy minimization problems that arise mainly in the context of computer vision tasks. This work motivates the use of such "hard-to-optimize" non-submodular functionals, and proposes methods and algorithms to cope with the NP-hardness of their optimization. Consequently, t…
Hard visual attention is a promising approach to reduce the computational burden of modern computer vision methodologies. Hard attention mechanisms are typically non-differentiable. They can be trained with reinforcement learning but the high-variance training this entails hinders more widespread application. We show h…
Attention operators have been widely applied in various fields, including computer vision, natural language processing, and network embedding learning. Attention operators on graph data enables learnable weights when aggregating information from neighboring nodes. However, graph attention operators (GAOs) consume exces…
Over recent years, devising classification algorithms that are robust to adversarial perturbations has emerged as a challenging problem. In particular, deep neural nets (DNNs) seem to be susceptible to small imperceptible changes over test instances. However, the line of work in provable robustness, so far, has been fo…
Autoregressive models struggle with hard-to-compute distributions, alternatives like energy-based and latent-variable models solve this.
problem Autoregressive models struggle with distributions whose next-symbol probability is hard to compute.
method Alternatives include energy-based models and latent-variable autoregressive models.
result Alternatives to autoregressive models can escape limitations of hard-to-compute distributions.
An algorithmically hard phase was described in a range of inference problems: even if the signal can be reconstructed with a small error from an information theoretic point of view, known algorithms fail unless the noise-to-signal ratio is sufficiently small. This hard phase is typically understood as a metastable bran…
Improved hardness results for clearing payments in financial networks with CDSs.
problem Determining clearing payments in financial networks with CDSs after financial shocks.
method Analyzing computational complexity of clearing problems, showing PPAD-hardness and FIXP-completeness improvements.
result PPAD-hardness of clearing problem significantly improved to ε ≈ 0.101.
Introduces a continuous version of LWE problem.
problem Hardness of learning mixtures of Gaussians.
method Polynomial-time quantum reduction from CLWE to lattice problems.
result CLWE shares hardness with LWE.
Training neural networks is hard in fixed dimensions.
problem Training two-layer neural networks is computationally hard in fixed dimensions.
method Parameterized complexity analysis considering dimension and number of neurons.
result Training two-layer neural networks is NP-hard for two dimensions.
Computing PL geometric category in 2D is NP-hard.
problem Determining the PL geometric category of 2D polyhedra.
method Reduction from shellability of 2-complexes, which is known to be NP-hard.
result It is NP-hard to decide whether the PL geometric category of a 2D polyhedron is at most 2.
Paper explores limits of high-order clustering with planted structures.
problem Statistical and computational limits of high-order clustering with planted structures.
method Developed methods for detection and recovery of clusters, identified signal-to-noise ratio boundaries.
result Sharp boundaries of signal-to-noise ratio for statistical and computational feasibility.
It is becoming increasingly important to understand the vulnerability of machine learning models to adversarial attacks. In this paper we study the feasibility of robust learning from the perspective of computational learning theory, considering both sample and computational complexity. In particular, our definition of…
Researchers prove quantum invariants remain hard even when restricted.
problem Computing quantum invariants on 3-manifolds with specific restrictions.
method Using Heegaard splittings and Hempel distance, they construct a hyperbolic 3-manifold with same invariant.
result Proving hardness of computing quantum invariants is preserved under specific restrictions.
New findings show learning deeper neural networks is hard even with Gaussian inputs and non-degenerate weights.
problem The computational complexity of learning neural networks, especially deeper ones.
method Smoothed analysis framework and local pseudorandom generators.
result Learning depth-3 ReLU networks under Gaussian input distribution is hard even if weight matrices are non-degenerate.
Model shows feature learning can improve neural scaling laws for hard tasks.
problem Understanding and improving neural network scaling laws for various task difficulties.
method Developed a solvable model of neural scaling laws, identified three scaling regimes, and demonstrated feature learning's impact on scaling exponents.
result Feature learning can improve scaling with training time and compute for hard tasks, nearly doubling the exponent.
New unsupervised method selects hard negative samples for contrastive learning.
problem How to select good negative examples for contrastive learning without using true similarity information.
method Developed a new family of unsupervised sampling methods for hard negative selection.
result Improves downstream performance across multiple modalities.
D-Wave computers struggle with sampling Boltzmann distributions efficiently.
problem Sampling Boltzmann distributions efficiently on D-Wave computers.
method Exploring various obstacles and remaining difficulties.
result Challenges remain in using D-Wave computers for efficient sampling.
Topological quantum computers use hyperbolic knots for computations.
problem The difficulty of calculating quantum invariants of knots.
method Using hyperbolic knots to compute topological quantum computer invariants.
result The hyperbolic geometry of knots is unlikely to be useful for topological quantum computation.
Studying general quantum many-body systems is one of the major challenges in modern physics because it requires an amount of computational resources that scales exponentially with the size of the system.Simulating the evolution of a state, or even storing its description, rapidly becomes intractable for exact classical…
We consider network sparsification as an L0-norm regularized binary optimization problem, where each unit of a neural network (e.g., weight, neuron, or channel, etc.) is attached with a stochastic binary gate, whose parameters are jointly optimized with original network parameters. The Augment-Reinforce-Merge (ARM),…
We show that {\sc Heegaard Genus ≤g}, the problem of deciding whether a triangulated 3-manifold admits a Heegaard splitting of genus less than or equal to g, is NP-hard. The result follows from a quadratic time reduction of the NP-complete problem {\sc CNF-SAT} to {\sc Heegaard Genus ≤g}.
Hard to approximate critical points for simple nonconvex functions.
problem Approximating critical points of nonconvex functions.
method Proving hardness results for polynomial-time approximation of critical points.
result Proving that approximating critical points is intractable for simple nonconvex functions.
We continue the study of statistical/computational tradeoffs in learning robust classifiers, following the recent work of Bubeck, Lee, Price and Razenshteyn who showed examples of classification tasks where (a) an efficient robust classifier exists, in the small-perturbation regime; (b) a non-robust classifier can be l…
There has recently been significant interest in hard attention models for tasks such as object recognition, visual captioning and speech recognition. Hard attention can offer benefits over soft attention such as decreased computational cost, but training hard attention models can be difficult because of the discrete la…
Paper shows hard computational limits for invariant causal prediction.
problem Hard computational limits for invariant causal prediction.
method Distributionally robust estimator with ellipse-shaped uncertain set.
result Estimation error rate can be arbitrarily slow for computationally efficient algorithms.
Quantum kernels offer potential speed-ups but require encoding problem-specific knowledge.
problem Generalization difficulty in high-dimensional feature spaces.
method Analysis of spectral properties of quantum kernels and their RKHS.
result Quantum advantage is expected if RKHS is low-dimensional and contains hard-to-compute functions.
The study constructs symplectic solvmanifolds satisfying the hard-Lefschetz condition.
problem Developing an analogue of Hodge theory for symplectic manifolds.
method Analyzing specific Lie algebras and their associated Lie groups, exploiting connections with Kneser graphs.
result Examples of almost-Kähler solvmanifolds satisfying the hard-Lefschetz condition are constructed.
TQFT invariants are either easy or hard to compute, depending on the TQFT type.
problem Computing TQFT invariants on closed 3-manifolds.
method Application of a dichotomy result for weighted constraint satisfaction problems over C.
result TQFT invariants are either solvable in polynomial time or #P-hard. Algorithm calculates quantum invariants of 3-manifolds with polynomial time complexity.
problem Computing quantum invariants from Tambara-Yamagami categories is #P-hard.
method Fixed-parameter tractable algorithm with first Betti number as parameter.
result Existence of FPT algorithm for Tambara-Yamagami invariants.
This paper uses QUBO to train machine learning models on quantum computers.
problem Efficiently training machine learning models on quantum computers.
method Formulated three machine learning models (linear regression, SVM, k-means) as QUBO problems.
result Formulations are more efficient or equivalent in time and space complexity to classical methods.
Trading system uses NP-hard optimization to select stocks for high Sharpe ratio trading.
problem Finding profitable, uncorrelated stocks for high Sharpe ratio trading.
method NP-hard combinatorial optimization using Ising machine and simulated bifurcation algorithm.
result Trading strategy with FPGA-based system achieves 164 μs response latency.
Study shortest non-separating curves on non-orientable surfaces, proving NP-hardness and tractability.
problem Computing shortest non-separating simple closed curves on non-orientable surfaces.
method Developed tools for computing shortest curves, proving NP-hardness and tractability.
result Proved NP-hardness and fixed-parameter tractability for computing shortest orienting curves, and polynomial-time algorithm for non-orienting curves.
We introduce a learning framework called learning using privileged information (LUPI) to the computer vision field. We focus on the prototypical computer vision problem of teaching computers to recognize objects in images. We want the computers to be able to learn faster at the expense of providing extra information du…
Optimal CL requires perfect memory and is NP-hard.
problem Designing CL algorithms that perform reliably and avoid catastrophic forgetting.
method Theoretical approach to derive computational properties of optimal CL algorithms.
result Optimal CL algorithms generally solve an NP-hard problem and require perfect memory.
New computational methods improve clustering of objects.
problem Clustering objects based on their joint occurrence in sets.
method Genetic algorithm with renumbering, local search, and simplified model.
result Improvements enhance computational performance of genetic algorithm.
New lower bounds for linear classification problems in high dimensions.
problem Linear classification problems in high-dimensional spaces.
method Reduction from hardness conjectures for Affine Degeneracy testing and k-Sum problems.
result Matching lower bounds of Ω(n^d) and respectively Ω(1/ε^d) for Maximum Halfspace Discrepancy problem.
Computer experiments reveal complex knots that don't simplify.
problem Understanding the dynamics of complex knots under self-repulsion.
method Computer simulations of knot theory, focusing on rational knots and tangles.
result Discovered hard unknots and complexified knots that do not reduce to simpler forms under self-repulsion.
Study shows optimal RL with transition look-ahead is NP-hard for ℓ≥2.
problem Optimal reinforcement learning with transition look-ahead is computationally hard.
method Proved NP-hardness for ℓ≥2 using linear programming. result There is a precise boundary between tractable and intractable cases for RL with look-ahead.
Learning shrinks hard tail, improving inference performance.
problem Improving inference performance in neural networks.
method Latent Instance Difficulty (LID) model analyzing fine-tuning of neural networks.
result Training-dependent inference scaling, with βexteff growing with sample size before saturating.