archive
Every paper Pith has read. Search by title, abstract, or pith.
14903 papers in cs.LG · page 15
-
Per-instance shapelets beat population averages on time-series tasks
INSHAPE: Instance-Level Shapelets for Interpretable Time-Series Classification
-
Evolutionary code agents gain by recycling deleted lines
What Do Evolutionary Coding Agents Evolve?
-
Adaptive Manifold Guidance conserves probability during strong guidance
Probability-Conserving Flow Guidance
-
Linear representation hypothesis enables efficient distillation to DP-aligned networks
Towards Distillation Guarantees under Algorithmic Alignment for Combinatorial Optimization
-
This paper proposes the Clipped Linear Lottery
Smooth Partial Lotteries for Stable Randomized Selection
-
Clipped Linear Lottery stabilizes randomized picks with near-optimal regret
Smooth Partial Lotteries for Stable Randomized Selection
-
Soft-log transform lets flow matching handle heavy tails
Tail Annealing for Heavy-Tailed Flow Matching
-
SDM improves adversarial attack performance and efficiency by reconstructing the…
SDM: A Powerful Tool for Evaluating Model Robustness
-
Active sampling achieves sqrt(n/T) ||p||_{2/3} regret bound
Active Context Selection Improves Simple Regret in Contextual Bandits
-
Critic disagreement guides reward poisoning in RIS networks
When Critics Disagree: Adaptive Reward Poisoning Attacks in RIS-Aided Wireless Control System
-
Diffusion model plans ride-hailing subsidies from past data alone
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
-
Diffusion planner lifts rides and GMV in ride-hailing under caps
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
-
Diffusion model raises ride-hailing rides and GMV under subsidy caps
D$^3$-Subsidy: Online and Sequential Driver Subsidy Decision-Making for Large-Scale Ride-Hailing Market
-
Mixture of experts catches text-camouflaged fraud in graphs
CAMERA: Adapting to Semantic Camouflage in Unsupervised Text-Attributed Graph Fraud Detection
-
Pointwise rejection costs replace global budgets in partial OT
Take It or Leave It: Intent-Controlled Partial Optimal Transport
-
Diffusion emulators make optimal particle filters training-free
Training-Free Bayesian Filtering with Generative Emulators
-
Two antagonistic Bayesian processes set the optimal learning rate
Training Neural Networks with Optimal Double-Bayesian Learning
-
Loss-adaptive rates cut forgetting by 93% in fine-tuning
Fine-Tuning Without Forgetting via Loss-Adaptive Learning Rates
-
LLM benchmarks can be made unlearnable to stop contamination
LLM Benchmark Datasets Should Be Contamination-Resistant
-
CryoNet maps debris-covered glaciers at 90 percent IoU
CryoNet: A Deep Learning Framework for Multi-Modal Debris-Covered Glacier Mapping. A Case Study of the Poiqu Basin, Central Himalaya
-
Four photodiodes replace cameras for robot odometry
Minimalist Visual Inertial Odometry
-
Visual encoder spatial detail fix unlocks precise robot tasks
Beyond Binary Success: A Diagnostic Meta-Evaluation Framework for Fine-Grained Manipulation
-
Lookahead training improves neural routing policies
Learning with Foresight: Enhancing Neural Routing Policy via Multi-Node Lookahead Prediction
-
Block-sphere quantizer lowers MSE and inner-product error
Block-Sphere Vector Quantization
-
Neighbors catch backdoors by sharing and comparing triggers
Your Neighbors Know: Leveraging Local Neighborhoods for Backdoor Detection in Decentralized Learning
-
Entropy change-point detection spots fluent LLM jailbreaks
Detecting Fluent Optimization-Based Adversarial Prompts via Sequential Entropy Changes
-
Local plasticity separates mixed signals by maximizing predictive entropy
Normative Networks for Source Separation via Local Plasticity and Dendritic Computation
-
Local rules recover sources from noisy mixtures
Normative Networks for Source Separation via Local Plasticity and Dendritic Computation
-
Rank-2 neural generators reach any orthonormal basis from Fourier
Learning Orthonormal Bases for Function Spaces
-
VLMs and agents miss over half the score on wild road damage
WildRoadBench: A Wild Aerial Road-Damage Grounding Benchmark for Vision-Language Models and Autonomous Agents
-
Non-negative weights make true DAG the unique minimizer
Exploiting Non-Negativity in DAG Structure Learning
-
Position-dependent attention fixes constant risk on shifted reasoning
A Measure-Theoretic Analysis of Reasoning: Structural Generalization and Approximation Limits
-
Gated estimator cuts manifold density error by 22-36%
Variance-Reduced Manifold Sampling via Polynomial-Maximization Density Estimation
-
Context map cache raises LLM agent accuracy 6-34% on recurring tasks
PEEK: Context Map as an Orientation Cache for Long-Context LLM Agents
-
Model fuses lidar and plot data for lower-bias forest biomass maps
StruMPL: Multi-task Dense Regression under Disjoint Partial Supervision and MNAR Labels
-
Parallel CFR runs 3.3 times faster on billion-history poker trees
Real-Time Parallel Counterfactual Regret Minimization
-
JAXenstein speeds first-person RL benchmarks several times
JAXenstein: Accelerated Benchmarking for First-Person Environments
-
Fast method learns node reps from labels without features
Fast and Featureless Node Representation Learning with Partial Pairwise Supervision
-
Contrastive learning separates local and global signals for multi-domain binding
Hierarchical Contrastive Learning for Multi-Domain Protein-Ligand Binding
-
Post-backprop rescaling fixes gradient scales in deep nets without BatchNorm
StableGrad: Backward Scale Control without Batch Normalization
-
Colluding accounts scale RAG privacy loss by sqrt(k)
Auditing Privacy in Multi-Tenant RAG under Account Collusion
-
Slice-wise tensorization yields near-lossless network compression
Fast Tensorization of Neural Networks via Slice-wise Feature Distillation
-
Framework fuses sensor data with physics rules for better passenger counts
A Closed-loop, State-centric, Multi-agent Framework for Passenger Load Estimation from Heterogeneous Data Streams
-
Diffusion model makes valid airfoils with lift and drag control
AirfoilGen: A valid-by-construction and performance-aware latent diffusion model for airfoil generation
-
Latent diffusion generates valid airfoils conditioned on lift and drag
AirfoilGen: A valid-by-construction and performance-aware latent diffusion model for airfoil generation
-
Set-valued policies output treatment sets instead of singles
Set-Valued Policy Learning
-
Domain cuts let neural operators handle PDE discontinuities
Smooth Piecewise Cutting for Neural Operator to Handle Discontinuities and Sharp Transitions
-
Explainer splits stable and changing links for temporal GNNs
ST-TGExplainer: Disentangling Stability and Transition Patterns for Temporal GNN Interpretability
-
Privacy budget sets floor on federated estimation error
General Lower Bounds for Differentially Private Federated Learning with Arbitrary Public-Transcript Interactions
2 Piths -
Benchmark separates ML models on flux extrapolation via tail errors
FLUXtrapolation: A benchmark on extrapolating ecosystem fluxes