archive
Every paper Pith has read. Search by title, abstract, or pith.
85951 papers indexed · page 148
-
GUI agents self-evolve via memory to hit 89% ScreenSpot
SE-GA: Memory-Augmented Self-Evolution for GUI Agents
-
Expert targets raise merged-model 4-bit accuracy from 35% to 77%
E-PMQ: Expert-Guided Post-Merge Quantization with Merged-Weight Anchoring
-
Multiple translations become one benchmark for Pali
PaliBench: A Multi-Reference Blueprint for Classical Language Translation Benchmarks
-
Dynamic graphs segment brain tumors with incomplete MRI data
Virtual Nodes Guided Dynamic Graph Neural Network for Brain Tumor Segmentation with Missing Modalities
-
Score-based diffusion recovers coherent masks for unseen manipulations
Towards Generalized Image Manipulation Localization via Score-based Model
-
Adversarial speech model improves asthma and COPD exacerbation detection
Speaker-Disentangled Remote Speech Detection of Asthma and COPD Exacerbations
-
Directional derivative measures faithfulness in zero-shot image explanations
Zero-Shot Faithful Textual Explanations via Directional-Derivative Influence on Predictions
-
Equation recovers spectral geometric mean for two operators
A multivariable mean equation arising from the spectral geometric mean
-
Statistical estimation reduces to stochastic optimization
Stochastic Optimization and Data Science
-
Correcting a few early tokens restores most LLM reasoning
Reasoning Can Be Restored by Correcting a Few Decision Tokens
-
Hallucination scores mask artifacts in diffusion 3D models
HAD: Hallucination-Aware Diffusion Priors for 3D Reconstruction
-
AI agents need a body to receive consequences or accountability fails
Some[Body] Must Receive That Pain for Agent Accountability
-
Subgoal conditioning boosts robot diffusion policy performance
SADP: Subgoal-Aware Diffusion Policy for Explainable Robots Learned from Foundation Model Generated Demonstrations
-
Proximal tendon loop cuts hysteresis error by 88 percent
SSTL: Self-Sensing Tendon Loop for Hysteresis Modeling and Compensation in Tendon-Sheath Mechanisms
-
Bounds restrict viable parameters for Type 3 gravity perturbations
Gauge-invariant cosmological perturbations in Type 3 New General Relativity and background-hierarchy bounds
-
Growing Jackson networks converge to infinite-dimensional fluid limits
Growing open Markovian Jackson networks: Fluid limit and infinite-dimensional Skorokhod problem
-
GoodServe lifts goodput up to 27.4% for agentic LLM inferences
GoodServe: Towards High-Goodput Serving of Agentic LLM Inferences over Heterogeneous Resources
-
Heavy tails make standard predictive tests unreliable
Heavy Tails and Predictive Ability Testing
-
DM test rejects true nulls up to 70% when errors have heavy tails
Heavy Tails and Predictive Ability Testing
-
MixSD retains 100% of base skills while injecting new facts
MixSD: Mixed Contextual Self-Distillation for Knowledge Injection
-
Mixing a model's own predictions lets it add facts without forgetting old skills
MixSD: Mixed Contextual Self-Distillation for Knowledge Injection
-
Label-free metrics fuse complementary VFM features for segmentation
Metric-Guided Feature Fusion of Visual Foundation Models for Segmentation Tasks
-
Graph plan computed first guides diffusion for long tasks
Plan First, Diffuse Later: Extrinsic Graph Guidance for Long-Horizon Diffusion Planning
-
Rigid wages and pegs cut inflation persistence
Wage Rigidity, Exchange Rate Regimes, and Inflation Persistence in Transition Economies: A Cohort-Based Institutional Approach
-
Causal prefix denoising lifts document recognition speed 71%
Prefix-Adaptive Block Diffusion for Efficient Document Recognition
-
Hybrid twin-LSTM cuts glucose forecast bias by 14 mg/dL at four hours
PhysioSeq2Seq: A Hybrid Physiological Digital Twin and Sequence-to-Sequence LSTM for Long-Horizon Glucose Forecasting in Type 1 Diabetes
-
Training-free method cuts 3D registration error up to 59%
VGGT-CD: Training-Free Robust Registration for 3D Change Detection
-
LLM agent reaches 96% success avoiding pedestrians in driving sim
Pedestrian-Aware LLM-Driven Behavioral Planning for Autonomous Vehicles
-
Adaptive memory boosts agent performance in multimodal tasks
Learning to Learn from Multimodal Experience
-
Nontrivial units vanish with high probability in random hypergraphs
Star-collision in random hypergraphs
-
Local guidance lifts LaCAM to higher lifelong MAPF throughput
Lifelong LaCAM with Local Guidance for Lifelong MAPF
-
Grading condition proves finite generation of cluster automorphism groups
Are cluster automorphism groups finitely generated?
-
Mechanism designs truthful social laws that maximize profit
A Truthful Multiunit Profit-Optimal Mechanism for Synthesizing Social Laws
-
Topfull graphs maximize 1-capacity via factorizations
Span capacities of graphs
-
Local Hölder continuity lifts to global for weighted α-subharmonic measure
Weighted $\alpha$-subharmonic measure
-
Wikipedia hyperlinks steer the order of edits
Editorial Trajectories in Wikipedia Reflect Underlying Hyperlink Structure
-
MoE models distributed across satellites run large AI on orbit
SpaceMoE: Towards Orbital General Intelligence with Distributed Mixture-of-Experts Inference
-
Induced patterns let VLMs plan beyond single-step vision
Thinking with Patterns: Breaking the Perceptual Bottleneck in Visual Planning via Pattern Induction
-
Multigraphs classify nonlinear equivariant operators on curved spaces
Equivariant nonlinear partial differential operators on constant curvature spaces
-
Two-layer wrapper makes model merges order-independent
Conflict-Free Replicated Data Types for Neural Network Model Merging: A Two-Layer Architecture Enabling CRDT-Compliant Model Merging Across 26 Strategies
-
New method cuts fractional-polynomial variance by closed-form factor
Efficient frequentist fractional polynomials for skewed dose-response and survival data: a variance-reducing alternative to OLS-FP
-
Flux residues emit visible light under VUV in xenon detectors
Photoluminescence of resin-based solder flux residue under ultraviolet excitation from 120 nm to 310 nm
-
AI needs named stage between narrow and general systems
Artificial Adaptive Intelligence: The Missing Stage Between Narrow and General Intelligence
-
First structured dataset released for Indian RTI decisions
RTI-Bench: A Structured Dataset for Indian Right-to-Information Decision Analysis
-
Staged RL lifts diffusion MLLM image quality
Sketch Then Paint: Hierarchical Reinforcement Learning for Diffusion Multi-Modal Large Language Models
-
THz silicon link delivers 1 Tbps aggregate throughput on chip
Dispersion-Engineered Terahertz Silicon Interconnects Enabling Terabit-Scale Data Links
-
No nontrivial t-edge-balanced graphs exist for t of 4 or more
On $t$-edge-balanced graphs
-
Block-union tables cut chunked prefill attention time by 2.72x
CompactAttention: Accelerating Chunked Prefill with Block-Union KV Selection
-
α-critical W_p graphs have three equivalent characterizations
A Ridge-Saturation Characterization of $\alpha$-Critical $\mathbf {W}_p$ Graphs
-
Hard thermodynamic rules inside AI produce consistent material models
Learning inelastic constitutive models from stress-strain data under hard thermodynamic constraints