PAC learning of networks from threshold opinion dynamics is efficient when influencers per agent are bounded but computationally hard for majority rules, with a heuristic succeeding in over 98% of simulations.
super hub Canonical reference
Overcoming catastrophic forgetting in neural networks
Canonical reference. 76% of citing Pith papers cite this work as background.
110
Pith papers citing it
4,193
external citations · OpenAlex
Background
76% of classified citations
hub tools
citation-role summary
background 26
method 2
baseline 1
citation-polarity summary
authors
co-cited works
representative citing papers
The pinwheel problem is NP-hard, implying hardness for related problems like bamboo garden trimming, and admits a PTAS that improves on the prior 9/7 approximation factor.
A new histogram-free likelihood method applied to simulated JWST observations of brown dwarfs shows that globular cluster ages can be determined with formal errors under 0.2 Gyr.
Proves non-Euclidean analogs of the Erdős-Anning theorem on integer-distance point sets and resolves Guy's equilateral dimension question via additively weighted Voronoi diagrams.
The Ramsey community number on the diamond hierarchical lattice is derived as an exact RG crossing of Bayesian evidence, with closed-form r_k and a thermodynamically ordered hierarchical community phase.
A framework integrating local direct reciprocity and global indirect reciprocity shows conditional cooperators resist invasion by unconditional strategies and that forgiving strategies best sustain cooperation.
Intention-use gaps and displacement of valued activities predict social media regret more strongly than duration, with pre-session context generalizing across users and physiological signals adding person-specific predictive power.
New triangular matrix representations and matrix-norm distances for comparing rooted ranked unlabeled phylogenetic networks.
Random and explicit (r,δ) quantum LRCs and h-level hierarchical QLRCs are constructed via CSS dual-containing codes, with distance bounds, a Singleton-like bound, and an efficient decoder for the Tamo–Barg family.
Introduces Abstraction Gap metric and CAGE benchmark showing seven of eight VLMs have large gaps between text plausibility and chain-based causal reasoning, with one model succeeding.
Introduces the 'innovation' property of LLMs and proves it is an almost characterization of hallucination while deriving new lower bounds on hallucination rates via missing mass.
Ultrametric graphons model hierarchical community networks and yield closed-form Laplacian spectra that approximate those of sampled random graphs with high probability as hierarchy depth grows.
Human face perception aligns with neural networks trained on inverse-generative and naturalistic discriminative tasks, as these best predict human dissimilarity judgments on controversial and random face pairs.
The PMNLV model extends single-neuron overdispersion to populations via matrix-normal gain priors, showing shared co-variability highest in V1 and declining along the mouse visual hierarchy.
Large-scale analysis of global popular music shows uncorrelated melodic and rhythmic diversities, with only rhythm linked to ethnic and linguistic heterogeneity.
A multilinear operator learned on PCA coefficients maps time-since-ignition inputs to smoke outputs, matching Monte Carlo accuracy with half the model calls and outperforming prior classifiers on holdout data.
Single-channel anonymization hides identity bias via cancellation effects, but full-pipeline anonymization reveals that homogeneous ensembles amplify sycophancy while heterogeneous ones reduce it, with one model showing unusually high baseline sycophancy.
TabDistill distills feature interactions from tabular foundation models via post-hoc attribution and inserts them into GAMs, yielding consistent predictive gains.
A geometric indicator from the normal width of the stochastic separatrix in a random two-state ecosystem model scales linearly with noise intensity and yields an affine relation to the logarithm of mean transition time.
Relaxed unitary convolutions for GNNs on meshes balance smoothness preservation with natural smoothing in dynamics, outperforming unitary convolutions and other models on PDEs and weather tasks.
An external magnetic field induces momentum-selective symmetry breaking in CsV3Sb5's electronic structure, consistent with piezomagnetism originating from vanadium Van Hove singularities at the charge density wave onset.
TPV measures first-order sensitivity of model outputs to parameter perturbations, unifies robustness analysis under one lens, proves train-to-test convergence in overparameterized limits, and enables label-free pruning and model selection applications.
Hybrid Bayesian-graph LLM agent reaches competitive performance against large models and achieves 67% win rate against humans in controlled Avalon play, outperforming baselines and human teammates.
DORI shows state-of-the-art multimodal models perform near-random on fine-grained object orientation reasoning despite strong scores on broader spatial benchmarks.
citing papers explorer
-
On the Limits of PAC Learning of Networks from Opinion Dynamics
PAC learning of networks from threshold opinion dynamics is efficient when influencers per agent are bounded but computationally hard for majority rules, with a heuristic succeeding in over 98% of simulations.
-
NP-Hardness and a PTAS for the Pinwheel Problem
The pinwheel problem is NP-hard, implying hardness for related problems like bamboo garden trimming, and admits a PTAS that improves on the prior 9/7 approximation factor.
-
New Way to Date Globular Clusters: Brown Dwarf Cooling Sequences
A new histogram-free likelihood method applied to simulated JWST observations of brown dwarfs shows that globular cluster ages can be determined with formal errors under 0.2 Gyr.
-
Non-Euclidean Erd\H{o}s-Anning Theorems
Proves non-Euclidean analogs of the Erdős-Anning theorem on integer-distance point sets and resolves Guy's equilateral dimension question via additively weighted Voronoi diagrams.
-
The Ramsey community number as a renormalization-group crossing
The Ramsey community number on the diamond hierarchical lattice is derived as an exact RG crossing of Bayesian evidence, with closed-form r_k and a thermodynamically ordered hierarchical community phase.
-
A model of local and global reciprocity
A framework integrating local direct reciprocity and global indirect reciprocity shows conditional cooperators resist invasion by unconditional strategies and that forgiving strategies best sustain cooperation.
-
Before You Scroll Again: Predicting Regretful Social Media Sessions from In-the-Wild Contextual and Wearable Sensing
Intention-use gaps and displacement of valued activities predict social media regret more strongly than duration, with pre-session context generalizing across users and physiological signals adding person-specific predictive power.
-
Matrix representations and distance metrics for unlabeled ranked phylogenetic networks
New triangular matrix representations and matrix-norm distances for comparing rooted ranked unlabeled phylogenetic networks.
-
Quantum Hierarchical Locally Recoverable Codes
Random and explicit (r,δ) quantum LRCs and h-level hierarchical QLRCs are constructed via CSS dual-containing codes, with distance bounds, a Singleton-like bound, and an efficient decoder for the Tamo–Barg family.
-
The Abstraction Gap in Vision-Language Causal Reasoning
Introduces Abstraction Gap metric and CAGE benchmark showing seven of eight VLMs have large gaps between text plausibility and chain-based causal reasoning, with one model succeeding.
-
Innovation: An Almost Characterization of Hallucination
Introduces the 'innovation' property of LLMs and proves it is an almost characterization of hallucination while deriving new lower bounds on hallucination rates via missing mass.
-
Ultrametric Graphons and Hierarchical Community Networks: Spectral Theory and Applications
Ultrametric graphons model hierarchical community networks and yield closed-form Laplacian spectra that approximate those of sampled random graphs with high probability as hierarchy depth grows.
-
Human face perception reflects inverse-generative and naturalistic discriminative objectives
Human face perception aligns with neural networks trained on inverse-generative and naturalistic discriminative tasks, as these best predict human dissimilarity judgments on controversial and random face pairs.
-
Partitioning Neural Co-Variability
The PMNLV model extends single-neuron overdispersion to populations via matrix-normal gain priors, showing shared co-variability highest in V1 and declining along the mouse visual hierarchy.
-
Do Melody and Rhythm Coevolve?
Large-scale analysis of global popular music shows uncorrelated melodic and rhythmic diversities, with only rhythm linked to ethnic and linguistic heterogeneity.
-
Enabling Real-Time Training of a Wildfire-to-Smoke Map with Multilinear Operators
A multilinear operator learned on PCA coefficients maps time-since-ignition inputs to smoke outputs, matching Monte Carlo accuracy with half the model calls and outperforming prior classifiers on holdout data.
-
Peer Identity Bias in Multi-Agent LLM Evaluation: An Empirical Study Using the TRUST Democratic Discourse Analysis Pipeline
Single-channel anonymization hides identity bias via cancellation effects, but full-pipeline anonymization reveals that homogeneous ensembles amplify sycophancy while heterogeneous ones reduce it, with one model showing unusually high baseline sycophancy.
-
Selecting Feature Interactions for Generalized Additive Models by Distilling Foundation Models
TabDistill distills feature interactions from tabular foundation models via post-hoc attribution and inserts them into GAMs, yielding consistent predictive gains.
-
Geometric early warning indicator from stochastic separatrix structure in a random two-state ecosystem model
A geometric indicator from the normal width of the stochastic separatrix in a random two-state ecosystem model scales linearly with noise intensity and yields an affine relation to the logarithm of mean transition time.
-
Smoothness Errors in Dynamics Models and How to Avoid Them
Relaxed unitary convolutions for GNNs on meshes balance smoothness preservation with natural smoothing in dynamics, outperforming unitary convolutions and other models on PDEs and weather tasks.
-
Magnetic field-induced momentum-dependent symmetry breaking in a kagome superconductor
An external magnetic field induces momentum-selective symmetry breaking in CsV3Sb5's electronic structure, consistent with piezomagnetism originating from vanadium Van Hove singularities at the charge density wave onset.
-
TPV: Parameter Perturbations Through the Lens of Test Prediction Variance
TPV measures first-order sensitivity of model outputs to parameter perturbations, unifies robustness analysis under one lens, proves train-to-test convergence in overparameterized limits, and enables label-free pruning and model selection applications.
-
Bayesian Social Deduction with Graph-Informed Language Models
Hybrid Bayesian-graph LLM agent reaches competitive performance against large models and achieves 67% win rate against humans in controlled Avalon play, outperforming baselines and human teammates.
-
Seeing Isn't Orienting: A Cognitively Informed Hierarchical Benchmark for Object Orientation in MLLMs
DORI shows state-of-the-art multimodal models perform near-random on fine-grained object orientation reasoning despite strong scores on broader spatial benchmarks.
-
A Paradigm Shift to Assembly-like Finite Element Model Updating
Assembly-like FEM updating performs ~95% of solves on subassemblies for 28% lower workload proxy effort with fidelity within 1% of global updating on experimental flexible wing data.
-
The Syncytial Mesh Model: A Mesoscale Control-Field Framework for Scale-Dependent Coherence in the Brain
The Syncytial Mesh Model proposes that astrocytic syncytial organization supplies a continuous mesoscale control field that shapes scale-dependent neuronal coherence and traveling-wave patterns beyond direct synaptic connectivity.
-
Gaming Consensus: Coordinated Manipulation in Crowdsourced Fact-Checking
Coordinated users can manipulate Community Notes' matrix factorization to make up to 10.7% of lower-quality notes appear helpful with fewer than 10 ratings, including a counterintuitive effect where 'Not Helpful' ratings can raise scores.
-
A Sieve-Accelerated Quadrature Method for Exact Privacy Accounting in the 2020 U.S. Decennial Census
A sieve-accelerated quadrature method using discrete Fourier transform enables the first exact privacy accounting for the 2020 Census DHC file under heterogeneous discrete Gaussian mechanisms with a claimed 1,824-fold speedup.
-
Correct codes for the wrong reasons? validating LLMs as measurement instruments for theoretical constructs
Grain calibration decomposes theoretical constructs into clause-level components, tests each with extractive evidence, and combines results through explicit theory-derived rules to validate LLM coding beyond agreement with human annotators.
-
The Effortless Trap: Productive Struggle, AI, and the Illusion of Learning
Proposes a six-move framework (Prime, Probe, Point, Attach, Strengthen, Test) for learning with AI, using an 'effortless' diagnostic to avoid illusion of mastery, backed by cited evidence of design-dependent outcomes including 17% harm from unguarded AI and doubled gains from engineered tutors.
-
Semantic Allocation in Ordered Bottlenecks: Predictive Residual Inference for Visual Representation Learning
PRIOR replaces masking-based ordering pressure with predictive residual inference using level-wise predictors to produce well-ordered representations that maintain or improve performance across budgets, especially in discrete settings.
-
Topological Out-of-Domain Generalization in Dynamical Systems Reconstruction
Proposes feature splitting and a closed-form bound on extrapolation range to enable zero-shot topological out-of-domain generalization in dynamical systems reconstruction across tipping points.
-
Super-earths and mini-neptunes follow different orbital period-eccentricity relations
Mini-neptunes exhibit an anti-correlation between orbital period and eccentricity unlike super-earths, implying the two populations are dynamically distinct.
-
The Geometry of Saturation: Effective Rank Predicts When Labels Stop Helping in Few-Shot Classification
The saturation index S(K)=erank(pooled within-class covariance)/K tracks when few-shot labels become redundant for fixed linear probes, with within-task median Spearman ρ≈0.81 on 17 binary tasks.
-
Mechanical Field Networks: Structured Neural Dynamics for Multivariate Systems
MF-Net learns a shared field state and mechanical transition rule from trajectories to deliver competitive forecasting and recoverable relation matrices on Lorenz-96 and real systems.
-
The Language of Elution: Autoregressive Prediction of the Next Feature in Untargeted LC-HRMS Lipidomics
Autoregressive LSTM and Transformer models achieve 98% top-1 accuracy predicting next eluting m/z bin from prior sequence features in lipidomics data across cohorts.
-
From Extrinsic to Intrinsic: Geodesic-Guided Representation Learning for 3D Geometric Data
PRISM is a pre-training method that learns isometric latent embeddings by explicitly recovering surface geodesic distances with a topology-enforcing loss and a two-stage training schedule.
-
Semigroup Consistency as a Diagnostic for Learned Physics Simulators
Normalized semigroup error is introduced as a diagnostic for learned simulators on 1D heat and Burgers equations; it correlates with rollout degradation (Spearman ρ=0.635) while regularization shows mixed results.
-
Two-Parameter Flows for Learning Population Dynamics of Physical Systems
Two-parameter flows learn base-to-marginal transports via conditional flow matching then extract unique physics-time velocities by regression on synthetic trajectories, inheriting regularity and scaling to high dimensions while permitting non-gradient dynamics.
-
Neuro-Inspired Inverse Learning for Planning and Control
The Inverter framework formalizes inverse learning to generate coherent multi-step trajectories, outperforming offline RL and diffusion baselines on D4RL maze tasks by 24% on average with 10-100x less inference time while also matching GRAPE fidelity on single-qubit gates at >1000x speed.
-
AMEL: Accumulated Message Effects on LLM Judgments
LLMs exhibit an accumulated message effect where conversation history polarity biases subsequent judgments, stronger for high-entropy items, independent of context length, and with a negativity bias.
-
Time-Dependent PDE-Constrained Optimization via Weak-Form Latent Dynamics
A WLaSDI-based framework creates noise-robust latent surrogates for PDE-constrained optimization, deriving direct and adjoint gradients to achieve up to five orders of magnitude speedup on radiative transfer, Vlasov-Poisson, and Burgers benchmarks.
-
The nestedness of higher-order networks
Unifies multiple nestedness measures in higher-order networks via the encapsulation DAG and demonstrates that nested structure is prevalent in social systems while its absence indicates mesoscale organization.
-
Geometric Phase Transition Enables Extreme Hippocampal Memory Capacity
A geometric phase transition produces crystalline hippocampal coding in food-caching birds that yields over 100-fold higher location memory capacity than the mist-like coding in non-caching birds.
-
Reasoning Can Be Restored by Correcting a Few Decision Tokens
Reasoning gaps between base LLMs and LRMs concentrate on ~8% of early planning tokens; intervening with the reasoning model only at high-disagreement positions recovers performance.
-
Will AI Agents Free Us From Meaningless Work? A Human-Centered Analysis
Task-level survey data shows perceived bullshitness predicts desire for AI delegation and reduced oversight needs.
-
Modeling Decision-Making with Will for Cooperation in Social Dilemmas
Willed agents that minimize distance to fixed goals shrink the feasible state space and catalyze cooperation in social dilemmas where continuous utility maximizers fail.
-
Discovering Ordinary Differential Equations with LLM-Based Qualitative and Quantitative Evaluation
DoLQ adds an LLM 'scientist' that judges candidate ODE terms both semantically and by numeric ablation, reporting the best structure recovery among the compared LLM-based ODE discovery methods.
-
Hall-Like Transversal Stress and Sandpile Criticality on Real Production Networks
The Hall-Sandpile model on WIOD networks generates four ordered regimes of instability where mean avalanche size and large-event probabilities increase with shock intensity and reduced redundancy, without evidence for universal power-law criticality.
-
Nothing Deceives Like Success: Social Learning and the Illusion of Understanding in Science
Success bias in collective theory-building leads to systematic overestimation of theory quality, narrower search, and paradoxically lower performance when agents optimize for apparent success.