archive
Every paper Pith has read. Search by title, abstract, or pith.
554 papers in cs.NE · page 8
-
The paper finds that a hierarchy-informed curricular evolutionary optimization (HICO) for…
Evolution With Purpose: Hierarchy-Informed Optimization of Whole-Brain Models
-
A two-time scale neurodynamic duplex based on three projection equations solves…
Distributionally Robust Geometric Joint Chance-Constrained Optimization: Neurodynamic Approaches
-
PC matches BP gradients exactly in wide residual networks
On the Infinite Width and Depth Limits of Predictive Coding Networks
-
Neuroevolution yields better chlorine controls for water networks
Optimizing Chlorination in Water Distribution Systems via Surrogate-assisted Neuroevolution
-
Energy-aware solvers match fitness using less energy
Energy-Aware Metaheuristics
-
Gradient descent races neurons to adapt network capacity
It's Not a Lottery, It's a Race: Understanding How Gradient Descent Adapts the Network's Capacity to the Task
-
Neural classifier and surrogate cut evaluations for Pareto search
NeuroPareto: Calibrated Acquisition for Costly Many-Goal Search in Vast Parameter Spaces
-
Backpropagation recovered from least-action principle
A Physical Theory of Backpropagation: Exact Gradients from the Least-Action Principle
-
Recurrent RL fine-tunes driving policies online with event cameras
Adaptive Control in Autonomous Driving via Real-Time Recurrent RL
-
MAR architecture cuts LLM energy use while matching dense accuracy
MAR: Efficient Large Language Models via Module-aware Architecture Refinement
-
Speciation keeps multiple toxic prompt niches alive in LLM search
Diversifying Toxicity Search in Large Language Models Through Speciation
-
Tournament task selection lifts quality and diversity in adversarial QD
Tournament Informed Adversarial Quality Diversity
-
MAP-Elites surfaces multiple strong motifs where MEME returns one
Motif Diversity in Human Liver ChIP-seq Data Using MAP-Elites
-
Metasurfaces turn wireless reception into an extreme learning machine
Over-The-Air Extreme Learning Machines with XL Reception via Nonlinear Cascaded Metasurfaces
-
Hypergraphs cut spike traffic in neuromorphic SNN mappings
A Case for Hypergraphs to Model and Map SNNs on Neuromorphic Hardware
-
Persistent assemblies evolve without genes or fitness rules
How Information Evolves: Stability-Driven Assembly and the Emergence of a Natural Genetic Algorithm
-
LSTM predicts ramp lane changes 4 seconds out at 76 percent accuracy
From Observation to Prediction: LSTM for Vehicle Lane Change Forecasting on Highway On/Off-Ramps
-
Evolutionary AI creates red-teaming systems hitting 100% ASR
AgenticRed: Evolving Agentic Systems for Red-Teaming
-
Quantum method lifts Sharpe ratio portfolios by 96.6 percent
A Quantum-Driven Evolutionary Framework for Solving High-Dimensional Sharpe Ratio Portfolio Optimization
-
DASH makes LLM solvers 4x faster for combinatorial problems
Rethinking LLM-Driven Heuristic Design: Generating Efficient and Specialized Solvers via Dynamics-Aware Optimization
-
Evolutionary library designs antennas near two-antenna array performance
ECLIPSE: An Evolutionary Computation Library for Instrumentation Prototyping in Scientific Engineering
-
Wachspress transfinite interpolants enforce exact Dirichlet boundaries in PINNs
A Wachspress-based transfinite formulation for exactly enforcing Dirichlet boundary conditions on convex polygonal domains in physics-informed neural networks
-
Model of discounts replaces histograms for high-dimensional QD
Discount Model Search for Quality Diversity Optimization in High-Dimensional Measure Spaces
-
QSLM framework slashes SLM memory by up to 86.5%
QSLM: A Performance- and Memory-aware Quantization Framework with Tiered Search Strategy for Spike-driven Language Models
-
Benchmark compares manual, deep learning, and foundation models for ERP
Benchmarking ERP Analysis: Manual Features, Deep Learning, and Foundation Models
-
Adaptive Markov thresholds keep eye-tracking accurate in noise
Identification of fixations and saccades in eye-tracking data using adaptive threshold-based method
-
Symmetric measure tracks net information change while conserving totals
Conserved active information
-
Ferroelectric synapses let SNNs adapt to individual EEG signals
Personalized Spiking Neural Networks with Ferroelectric Synapses for EEG Signal Processing
-
Symbolic regression finds equations in physical data
Introduction to Symbolic Regression in the Physical Sciences
-
Depth selection makes KAN models smaller and accurate
Optimized Architectures for Kolmogorov-Arnold Networks
-
Asynchronous search widens LLM accuracy-cost Pareto fronts
AP-BMM: Approximating Capability-Cost Pareto Sets of LLMs via Asynchronous Prior-Guided Bayesian Model Merging
-
Equivariance cuts neural net expressivity but bigger models fix it
Drawback of Enforcing Equivariance and its Compensation via the Lens of Expressive Power
-
Local build lets spiking networks scale to thousands of GPUs
Scalable Construction of Spiking Neural Networks using up to thousands of GPUs
-
Brush turns patient data into simple clinical rules
Towards symbolic regression for interpretable clinical decision scores
-
Dual-memory spiking nets cut parameters 40-60% while raising throughput 4X
Algorithm-hardware co-design of neuromorphic networks with dual memory pathways
-
GraphBench unifies benchmarks for graph learning models
GraphBench: Next-generation graph learning benchmarking
-
Search finds energy-saving spiking nets for eye emotion recognition
Neural Architecture Search of Time-to-First-Spike-Coded Spiking Neural Networks for Efficient Eye-based Emotion Recognition
-
Neuromorphic system detects falls at 84% F1 using 90 mW
Privacy-preserving fall detection at the edge using Sony IMX636 event-based vision sensor and Intel Loihi 2 neuromorphic processor
-
ARRDE holds performance steady across five varied optimization suites
Robust Differential Evolution via Nonlinear Population Size Reduction and Adaptive Restart: The ARRDE Algorithm
-
ARRDE keeps DE stable across five CEC suites
Robust Differential Evolution via Nonlinear Population Size Reduction and Adaptive Restart: The ARRDE Algorithm
-
ANNs replicate brain navigation cells to cut path-integration cost
Boosting Brain-inspired Path Integration Efficiency via Learning-based Replication of Continuous Attractor Neurodynamics
-
Rare jumps explain most variance in volume scale
Jump-diffusion models of parametric volume-price distributions
-
Multi-agent optimization refines Chinese baby names with explanation checks
Chinese Short-Form Creative Content Generation via Explanation-Oriented Multi-Objective Optimization
-
Kernel Hopfield networks form spectral ridge for stability and capacity
Self-Organization and Spectral Mechanism of Attractor Landscapes in High-Capacity Kernel Hopfield Networks
-
Single transformer guides genetic programming to top-ranked symbolic regressions
Transformer Semantic Genetic Programming for d-dimensional Symbolic Regression Problems
-
Clustering pixels in space and time lets SNNs classify digits with 24 percent fewer spikes
Spatio-Temporal Cluster-Triggered Encoding for Spiking Neural Networks
-
Dual ranking on fitness and uncertainty improves offline MOO
Uncertainty-Aware Offline Data-Driven Multi-Objective Optimization
-
Quantization cuts memory fourfold for edge AI models
Quantization Effects of Artificial Neural Networks for Embedded Edge-Computing Applications
-
Internal causal transmission across levels generates consciousness
Considering a generative mechanism of consciousness from the perspective of inter-level causation
-
Evolutionary optimizer beats Adam for diffusion prompt search
Evolutionary Optimization Trumps Adam Optimization on Embedding Space Exploration