p-MEM stores distribution parameters and samples at native memory bandwidth, achieving over 1000 GSa/s/mm² GRNG throughput and reducing sampling latency and energy in probabilistic edge workloads.
hub
Title resolution pending
12 Pith papers cite this work. Polarity classification is still indexing.
12
Pith papers citing it
hub tools
representative citing papers
A transformer-encoded spherical normalizing flow achieves state-of-the-art angular resolution for IceCube neutrino tracks and showers, improving median resolution by factors of 1.3-2.5 over B-spline likelihoods at 100 TeV and outperforming prior ML methods for muons.
Tabular diffusion models leak membership information via attacks even with partial attacker knowledge, and common heuristic privacy metrics like distance-to-closest-record are unreliable.
Multi-stage LLM pipeline creates validated BPMN models from text and reconstructs them with average similarity above 0.75 across 387 cases from 750 public diagrams.
Truncated kernel SGD with spherical RBFs projects stochastic gradients to a finite hypothesis space for minimax-optimal rates on optimization, generalization, and strong RKHS convergence across losses like least-squares and logistic.
A supervisor-specialist multi-agent system with structured artifact reuse and dynamic replanning improves planning effectiveness by 54.5% and task completion by 37.8% over baseline for industrial maintenance QA while cutting tool time share from 47.3% to 26.3%.
AML outperforms cross-validated baselines including CNNs on 50-2000 example image datasets and is comparable to XGBoost/LightGBM on tabular data using only training data and no task-dependent hyperparameters.
A three-layer probabilistic assume-guarantee architecture is structurally required for safe LLM agent deployment.
Sliding-window bidirectional-attention KV compression with Token2Chunk and periodic PagedAttention eviction keeps near-full reasoning accuracy at ~20% cache while lifting memory-constrained throughput ~12.75%.
Prompt-driven image-to-video generation produces deictic gestures that match real data visually, add useful variety, and improve downstream recognition models when mixed with human recordings.
Causal Path Alignment anchors optimization trajectories in in-parameter knowledge editing to follow relation-aware causal paths, reducing subject-dominant memory interference in LLMs.
citing papers explorer
-
Probabilistic Memory for Trustworthy Edge Intelligence
p-MEM stores distribution parameters and samples at native memory bandwidth, achieving over 1000 GSa/s/mm² GRNG throughput and reducing sampling latency and energy in probabilistic edge workloads.
-
Neural posterior estimation of the neutrino direction in IceCube using transformer-encoded normalizing flows on the sphere
A transformer-encoded spherical normalizing flow achieves state-of-the-art angular resolution for IceCube neutrino tracks and showers, improving median resolution by factors of 1.3-2.5 over B-spline likelihoods at 100 TeV and outperforming prior ML methods for muons.
-
On Privacy Leakage in Tabular Diffusion Models: Influential Factors, Attacker Knowledge, and Metrics
Tabular diffusion models leak membership information via attacks even with partial attacker knowledge, and common heuristic privacy metrics like distance-to-closest-record are unreliable.
-
Automated BPMN Model Generation from Textual Process Descriptions: A Multi-Stage LLM-Driven Approach
Multi-stage LLM pipeline creates validated BPMN models from text and reconstructs them with average similarity above 0.75 across 387 cases from 750 public diagrams.
-
Truncated Kernel Stochastic Gradient Descent with General Losses and Spherical Radial Basis Functions
Truncated kernel SGD with spherical RBFs projects stochastic gradients to a finite hypothesis space for minimax-optimal rates on optimization, generalization, and strong RKHS convergence across losses like least-squares and logistic.
-
Towards Multi-Turn Dialog Systems for Industrial Asset Operations and Maintenance
A supervisor-specialist multi-agent system with structured artifact reuse and dynamic replanning improves planning effectiveness by 54.5% and task completion by 37.8% over baseline for industrial maintenance QA while cutting tool time share from 47.3% to 26.3%.
-
Algebraic Machine Learning for Small-to-Medium Datasets Is Competitive against Strong Standard Baselines
AML outperforms cross-validated baselines including CNNs on 50-2000 example image datasets and is comparable to XGBoost/LightGBM on tabular data using only training data and no task-dependent hyperparameters.
-
Position: A Three-Layer Probabilistic Assume-Guarantee Architecture Is Structurally Required for Safe LLM Agent Deployment
A three-layer probabilistic assume-guarantee architecture is structurally required for safe LLM agent deployment.
-
KARA: Efficient Reasoning LLM Serving via Sliding-Window KV Cache Compression
Sliding-window bidirectional-attention KV compression with Token2Chunk and periodic PagedAttention eviction keeps near-full reasoning accuracy at ~20% cache while lifting memory-constrained throughput ~12.75%.
-
Prompt-to-Gesture: Measuring the Capabilities of Image-to-Video Deictic Gesture Generation
Prompt-driven image-to-video generation produces deictic gestures that match real data visually, add useful variety, and improve downstream recognition models when mixed with human recordings.
-
Causal Path Alignment: Anchoring the Optimization Trajectory for Controllable In-Parameter Knowledge Editing
Causal Path Alignment anchors optimization trajectories in in-parameter knowledge editing to follow relation-aware causal paths, reducing subject-dominant memory interference in LLMs.
- The CTI Echo Chamber: Fragmentation, Overlap, and Vendor Specificity in Twenty Years of Cyber Threat Reporting