archive
Every paper Pith has read. Search by title, abstract, or pith.
55131 papers indexed · page 15
-
Benchmark standardizes early Parkinson's speech detection
A Benchmark for Early-stage Parkinson's Disease Detection from Speech
2 Piths -
Rodrigues integrator rivals Boris for turbulence particle tracking
Boris and Exponential Integrators in the Theory of Particles Interacting with Magnetic Turbulence
-
Arrest and instabilities compete to form varied amorphous states
Interference of dynamical arrest, thermodynamic instabilities and energy-scale competition in symmetric binary mixtures
-
Entropy gate turns off source anchor when frozen model degrades
Reliability-Gated Source Anchoring for Continual Test-Time Adaptation
-
Early rejection cuts LLM synthetic data tokens by 11-77%
Know When To Fold 'Em: Token-Efficient LLM Synthetic Data Generation via Multi-Stage In-Flight Rejection
-
MathAtlas benchmark shows models reach under 10 percent on graduate math
MathAtlas: A Benchmark for Autoformalization in the Wild
2 Piths -
Penalized heat controls converge to hard terminal states at rate O(1/alpha)
Quantitative Soft-to-Hard Terminal Constraint Convergence for the Heat Equation
-
Ising networks store continuous signals near 0.5 capacity
Finite-size scaling of hetero-associative retrieval in continuous-signal-driven Ising spin systems
-
Non-convex solvers find exact nonnegative ranks for some matrices
Computing Lower Bounds on the Nonnegative Rank via Non-Convex Optimization Solvers
-
Dual RL agents learn to probe like Supreme Court justices
Dual Hierarchical Dialogue Policy Learning for Legal Inquisitive Conversational Agents
-
MCMC method gives reliable uncertainty for fMRI brain connectivity
An MCMC-Based Method for Dynamic Causal Modeling of Effective Connectivity in Functional MRI
-
New method lifts multi-task LLM accuracy by 6.67 percent
PEML: Parameter-efficient Multi-Task Learning with Optimized Continuous Prompts
-
Separate rewards for seeing and thinking lift VLM results on both
Bad Seeing or Bad Thinking? Rewarding Perception for Vision-Language Reasoning
-
Logic rules in prompts cut RAG errors via derivation trees
Derivation Prompting: A Logic-Based Method for Improving Retrieval-Augmented Generation
-
Biharmonic maps to spheres satisfy energy identity for n at least 5
Energy identity for stationary biharmonic mappings into spheres in supercritical dimensions
-
SPIN cuts executed tasks from 1061 to 623 on industrial benchmarks
SPIN: Structural LLM Planning via Iterative Navigation for Industrial Tasks
-
Gravitons reflected via photon conversion for lab lasers
Reflecting Gravitons: The Graviton Laser and the Gertsenshtein effect
-
Formal checks can keep AI legal reasoning inside the text
Bridging Legal Interpretation and Formal Logic: Faithfulness, Assumption, and the Future of AI Legal Reasoning
-
Network-aware tokenization stabilizes brain connectivity learning
Network-Aware Bilinear Tokenization for Brain Functional Connectivity Representation Learning
-
ViTs reach 84% accuracy by replacing layer norm with evolved scalars
Evolving Layer-Specific Scalar Functions for Hardware-Aware Transformer Adaptation
-
Galois actions give non-special divisors for arbitrary Kummer ramification
Construction of Non-special Divisors on Kummer Covers With Arbritary Ramification For LCP Codes
-
VLM perceptions guide rectified flow to clear mixed weather effects
PVRF: All-in-one Adverse Weather Removal via Prior-modulated and Velocity-constrained Rectified Flow
-
Magnetic fields create ultralight black holes in inflation
The Magnetic Origin of Primordial Black Holes: Ultralight PBHs and Secondary GWs
-
RL embeds solar-battery sizes into multi-market bidding policy
Optimal design of solar-battery hybrid resources considering multi-market participation under weather and price uncertainty
-
Compiler cuts fault-tolerant quantum runtimes by up to 60x
C-Phase-Aware Compilation for Efficient Fault-Tolerant Quantum Execution
-
Penalized deep networks equal MAP estimates under hierarchical priors
Wahkon: A Statistically Principled Deep RKHS Superposition Network
-
Audited data lifts 8B model 18 points on physics olympiads
Physics-R1: An Audited Olympiad Corpus and Recipe for Visual Physics Reasoning
-
Instantaneous frequency fitting beats Nyquist limit in FMCW lidar
FMCW Lidar Beyond Nyquist by Instantaneous Frequency Fitting
-
LLMs know when to use tools but often skip the call
Model-Adaptive Tool Necessity Reveals the Knowing-Doing Gap in LLM Tool Use
-
Learned predictor prunes KV cache 3-10x on the fly
Self-Pruned Key-Value Attention: Learning When to Write by Predicting Future Utility
-
Unary recoding enables polynomial-time rule learning for LLMs
Enhanced and Efficient Reasoning in Large Learning Models
-
The ℓFEM package offers a loop-free MATLAB implementation of isoparametric finite…
$\ell$FEM: An efficient loop-free Matlab implementation of isoparametric bulk and surface finite elements
-
GraphRAG retrieval aligns LLM agents with social values
From Descriptive to Prescriptive: Uncover the Social Value Alignment of LLM-based Agents
-
Obstruction ranks correct theory shifts for AI agents
Sheaf-Theoretic Transport and Obstruction for Detecting Scientific Theory Shift in AI Agents
-
Fingerprinting blocks RRC storms in 5G O-RAN before exhaustion
StormShield: Fingerprint-Based Detection and Mitigation of RRC Signaling Storms in O-RAN 5G RANs
-
General audio pretraining beats domain MAE for bioacoustics
Masked Autoencoders with Limited Data: Does It Work? A Fine-Grained Bioacoustics Case Study
-
Billiard complexity computed exactly for hyperbolic polygons with even q
Complexity of Billiards in Polygons Associated to Hyperbolic $(p,q)$-Tilings
-
Multi-term error metric simplifies noisy meshes faster
Fast and Robust Mesh Simplification for Generated and Real-World 3D Assets
-
Each pixel receives its own token in new language model
Unified Pix Token And Word Token Generative Language Model
-
No-go theorem blocks static wormholes via vacuum coupling
A No-Go Theorem for Topological Bridges with Matter-Vacuum Coupling
-
Non-centered objective stabilizes SPL representations at high UTD
R2R2: Robust Representation for Intensive Experience Reuse via Redundancy Reduction in Self-Predictive Learning
-
Controls explain all apparent language model to brain alignment
Do Language Models Align with Brains? Prediction Scores Are Not Enough
-
Variational method simplifies solitary waves for extended KdV
A study of variational single solitary waves governed by the conservative-extended KdV equation with applications to shallow water dispersive shocks
-
Bent functions give optimal parity for Hamming membership checks
Function-Correction with Optimal Data Protection for the General Hamming Code Membership
-
Flow control cuts drag 49% with 37x less energy
Policy-DRIFT: Dynamic Reward-Informed Flow Trajectory Steering
-
AI Overviews appear in 14% of searches with 11% unsupported claims
Measuring Google AI Overviews: Activation, Source Quality, Claim Fidelity, and Publisher Impact
-
Memory forensics framework extracts runtime state from Go malware
Memory Forensics Techniques for Automated Detection and Analysis of Go Malware
-
Regret equals covariance between costs and decisions
Regret Equals Covariance: A Closed-Form Characterization for Stochastic Optimization
-
ET Cru secondary shows deep CNO-cycle exposure
Spectral Disentangling Reveals Deep CNO-cycle Exposure in ET Cru
-
MICADO on ELT to constrain dark cluster around Sgr A*
Constraining the Galactic Center Dark Cluster with ELT/MICADO Observations