AI reviews for all 22,977 AAAI-26 papers were preferred by authors and PC members over human reviews on accuracy and suggestions and outperformed baselines at spotting weaknesses.
Canonical reference
Title resolution pending
Canonical reference. 80% of citing Pith papers cite this work as background.
59
Pith papers citing it
Background
80% of classified citations
citation-role summary
background 15
citation-polarity summary
fields
cs.HC 7 cs.LG 7 cs.CL 5 q-bio.NC 5 cs.AI 4 cs.CV 3 q-bio.PE 3 cs.CY 2 cs.DL 2 physics.soc-ph 2roles
background 15representative citing papers
A Fourier model of the fly compound eye shows zebra stripes generate parasitic Moiré frequencies that cancel radial expansion signals and disrupt landing at 1–5 m.
The authors synthesize a typology of fourteen OSS sub-genres from a review of 3,925 papers and present a research agenda on cross-sub-genre generalization.
A taxonomy of SNN training algorithms is presented with the release of NeuroTrain, an open benchmarking framework for reproducible comparisons across datasets and architectures.
A 14-code content model for local post-hoc AI explanations, derived from 325 user statements and validated by experts with high reliability scores.
The ultrametric phylogenetic Laplacian has closed-form eigenvalues that aggregate clade-weighted branch lengths and eigenvectors supported on individual clades, enabling linear-time spectral reconstruction and eigenmode analysis of traits.
In-weights learning induces linear embeddings enabling transitive inference in transformers, whereas in-context learning defaults to match-and-copy unless pre-trained on linear tasks or prompted with linear mental maps.
DLCM is a lattice-based stochastic solver for multicellular systems that introduces elliptic projection to couple cell pressure with population curvature for surface tension effects.
Introduces a quantum-analogue cloud-function model for large-scale sensory processing in the brain and uses it to account for post-decisional changes of mind via interplay of fast and slow neural dynamics.
Data points genuinely similar to multiple dissimilar neighborhoods are detected as local articulation points of a sparsified high-dimensional graph and split into multiple projected copies, one per neighborhood.
KAPLAN-HR applies B-spline KANs to nonparametric hazard estimation in survival analysis, recovering GAMs in the single-layer case, capturing interactions via deeper layers, with convergence rates independent of covariate dimension for KAN-representable targets, and competitive performance on six cli
The sTM model is extended with sequential population activations to encode element durations across timescales and uses oscillatory inputs as a clock signal to modulate replay speed in spiking networks.
PULSE demonstrates that agentic LLM-based investigation of passive smartphone sensing data achieves balanced accuracies of 0.743 (with diary) and 0.713 (sensing-only) for predicting emotion regulation desire and intervention availability in 50 cancer survivors.
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.
CORE-Cox learns low-rank Cox coefficients across outcomes in a source cohort then applies regularized adaptation to a target cohort, yielding C-index gains from 0.733 to 0.766 in UK Biobank and 0.628 to 0.658 in MIMIC-IV Asian subgroups under nested cross-validation.
TechToken uses transformer embeddings of IPC codes to measure linguistic convergence in patents and predict future technological combinations.
Speculative design societies prompted 12 OSS practitioners to reflect critically on why designers are underrepresented and to suggest actions for more inclusive open source environments.
Hybrid models that add persistent-homology features from fixation time series to traditional statistical features outperform purely statistical baselines for dyslexia detection on the Copenhagen Corpus.
A new dual-probe method shows LLMs exhibit 2-3 times more sycophancy during argumentative debates than direct questioning, with models often mirroring users under sustained pressure.
A modular two-stage selective rescanning framework for SEM achieves 99% detection of damage features in DP800 steel at 58% of conventional acquisition time.
A single spatial correlation index predicts cooperator fixation across arbitrary heterogeneous landscapes, with segregated environments enhancing and intermixed ones suppressing cooperation under weak selection.
New imbalanced NGS dataset pairs QC-34 features with ENCODE blocklist features on the same human and mouse samples to study quality control prediction.
A smartwatch system detected social interactions in daily life with 77% of its alerts confirmed by users, and a multi-sensor model reached 90% balanced accuracy on 15-second windows.
Agent-based AI workflows repair injected reproducibility failures in R social-science code at 69-96% success, substantially outperforming prompt-based LLM approaches at 31-79%.
citing papers explorer
-
AI-Assisted Peer Review at Scale: The AAAI-26 AI Review Pilot
AI reviews for all 22,977 AAAI-26 papers were preferred by authors and PC members over human reviews on accuracy and suggestions and outperformed baselines at spotting weaknesses.
-
Biting fly vision and zebra stripes
A Fourier model of the fly compound eye shows zebra stripes generate parasitic Moiré frequencies that cancel radial expansion signals and disrupt landing at 1–5 m.
-
Open Source Is Not One Thing: A Typology of Open-Source Software Sub-Genres
The authors synthesize a typology of fourteen OSS sub-genres from a review of 3,925 papers and present a research agenda on cross-sub-genre generalization.
-
NeuroTrain: Surveying Local Learning Rules for Spiking Neural Networks with an Open Benchmarking Framework
A taxonomy of SNN training algorithms is presented with the release of NeuroTrain, an open benchmarking framework for reproducible comparisons across datasets and architectures.
-
What Should Explanations Contain? A Human-Centered Explanation Content Model for Local, Post-Hoc Explanations
A 14-code content model for local post-hoc AI explanations, derived from 325 user statements and validated by experts with high reliability scores.
-
Spectral Geometry and Heat Kernels on Phylogenetic Trees
The ultrametric phylogenetic Laplacian has closed-form eigenvalues that aggregate clade-weighted branch lengths and eigenvectors supported on individual clades, enabling linear-time spectral reconstruction and eigenmode analysis of traits.
-
Relational reasoning and inductive bias in transformers and large language models
In-weights learning induces linear embeddings enabling transitive inference in transformers, whereas in-context learning defaults to match-and-copy unless pre-trained on linear tasks or prompted with linear mental maps.
-
DLCM: a versatile multi-level solver for heterogeneous multicellular systems
DLCM is a lattice-based stochastic solver for multicellular systems that introduces elliptic projection to couple cell pressure with population curvature for surface tension effects.
-
A Quantum-Analogue Formalism for Modeling Supraliminal Information Processing
Introduces a quantum-analogue cloud-function model for large-scale sensory processing in the brain and uses it to account for post-decisional changes of mind via interplay of fast and slow neural dynamics.
-
When One Point Is Not Enough: Addressing Ambiguous Instances in Dimensionality Reduction by Splitting
Data points genuinely similar to multiple dissimilar neighborhoods are detected as local articulation points of a sparsified high-dimensional graph and split into multiple projected copies, one per neighborhood.
-
KAPLAN: Kolmogorov-Arnold Prognostic Learnable Activation Networks for Survival Analysis
KAPLAN-HR applies B-spline KANs to nonparametric hazard estimation in survival analysis, recovering GAMs in the single-layer case, capturing interactions via deeper layers, with convergence rates independent of covariate dimension for KAN-representable targets, and competitive performance on six cli
-
Learning sequence timing and control of replay speed in networks of spiking neurons
The sTM model is extended with sequential population activations to encode element durations across timescales and uses oscillatory inputs as a clock signal to modulate replay speed in spiking networks.
-
PULSE: Agentic Investigation with Passive Sensing for Proactive Intervention in Cancer Survivorship
PULSE demonstrates that agentic LLM-based investigation of passive smartphone sensing data achieves balanced accuracies of 0.743 (with diary) and 0.713 (sensing-only) for predicting emotion regulation desire and intervention availability in 50 cancer survivors.
-
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.
-
Structured Transfer Learning for Survival Risk Stratification in Data-Sparse Clinical Cohorts
CORE-Cox learns low-rank Cox coefficients across outcomes in a source cohort then applies regularized adaptation to a target cohort, yielding C-index gains from 0.733 to 0.766 in UK Biobank and 0.628 to 0.658 in MIMIC-IV Asian subgroups under nested cross-validation.
-
Anticipating Innovation Using Large Language Models
TechToken uses transformer embeddings of IPC codes to measure linguistic convergence in patents and predict future technological combinations.
-
What If We Work Together? Fostering Reflections on Designer Inclusion in Open Source Software Through Speculative Design
Speculative design societies prompted 12 OSS practitioners to reflect critically on why designers are underrepresented and to suggest actions for more inclusive open source environments.
-
Fixation Sequences as Time Series: A Topological Approach to Dyslexia Detection
Hybrid models that add persistent-homology features from fixation time series to traditional statistical features outperform purely statistical baselines for dyslexia detection on the Copenhagen Corpus.
-
Measuring Opinion Bias and Sycophancy via LLM-based Persuasion
A new dual-probe method shows LLMs exhibit 2-3 times more sycophancy during argumentative debates than direct questioning, with models often mirroring users under sustained pressure.
-
SPARSE -- Efficient High-Resolution SEM Imaging of Rare Microstructural Features Across Large Areas by Selective Rescanning
A modular two-stage selective rescanning framework for SEM achieves 99% detection of damage features in DP800 steel at 58% of conventional acquisition time.
-
Predicting success of cooperators across arbitrary heterogeneous environmental landscapes
A single spatial correlation index predicts cooperator fixation across arbitrary heterogeneous landscapes, with segregated environments enhancing and intermixed ones suppressing cooperation under weak selection.
-
An Imbalanced Dataset with Multiple Feature Representations for Studying Quality Control of Next-Generation Sequencing
New imbalanced NGS dataset pairs QC-34 features with ENCODE blocklist features on the same human and mouse samples to study quality control prediction.
-
SocialPulse: On-Device Detection of Social Interactions in Naturalistic Settings Using Smartwatch Sensing
A smartwatch system detected social interactions in daily life with 77% of its alerts confirmed by users, and a multi-sensor model reached 90% balanced accuracy on 15-second windows.
-
Automating Computational Reproducibility in Social Science: Comparing Prompt-Based and Agent-Based Approaches
Agent-based AI workflows repair injected reproducibility failures in R social-science code at 69-96% success, substantially outperforming prompt-based LLM approaches at 31-79%.
-
PALMS: A Computational Implementation for Pavlovian Associative Learning Models' Simulation
PALMS is a computational tool implementing canonical and attentional Pavlovian learning models with support for large experiments and a new unified learning rate variant that combines Mackintosh and Pearce-Hall ideas.
-
Detecting In-Person Conversations in Noisy Real-World Environments with Smartwatch Audio and Motion Sensing
A multimodal machine learning framework fusing smartwatch audio and inertial sensing achieves macro F1 scores of 82% in lab and 77% in semi-naturalistic studies for detecting face-to-face conversations.
-
Efficient stochastic simulation of gene regulatory networks using hybrid models of transcriptional bursting
A new algorithm simulates exact trajectories of bursty PDMP models for arbitrary interacting genes at lower cost than SSA, with a toggle-switch example showing bimodal expression arises from interaction-driven burst frequencies rather than bursting alone.
-
Unified Framework for Binary-Choice Dynamics: Analysis and Applications
When transition rates of two competing mechanisms sum equally, annealed and quenched binary-choice dynamics coincide, heterogeneity reduces to the mean preference, and oscillations are impossible.
-
May (A)I Beautify Your Visualization? Expert Judgments of Acceptable Aesthetic Alterations
Expert survey finds acceptability of visualization alterations driven by transformation meaning, with AI changes rated less acceptable than identical human ones.
-
Multi-Level Distributional Entropy for Explainable Network Intrusion Detection
MDE computes three entropy features from flow stats to match conventional ML performance (F1 0.708-0.989) on four IDS benchmarks while exposing aggregate-metric failures and providing stable SHAP attributions.
-
Quantifying the Salience of Geo-Cultural Values for Pluralistic Safety Alignment
Cultural zones explain variance in safety ratings beyond demographics across six datasets, with roughly 10% of items identified as culturally sensitive.
-
Experimental Implementation of the Quantum Volunteer's Dilemma on NISQ Hardware: Noise Analysis and Digital-Twin Validation
Experimental demonstration of multiplayer quantum volunteer's dilemma on NISQ hardware up to 9 players, with noise analysis and digital-twin comparison.
-
Audio Deepfake Detection with Half-Truth Localisation Using Cross-Attentive Feature Fusion
CAFNet performs joint ternary classification and temporal boundary regression for half-truth audio deepfakes via cross-attentive fusion of MFCC, LFCC, and Chroma-STFT features, reporting 92.71% accuracy and 0.075s MAE on MLADDC T2+T3.
-
Copula and spatial-regularized variational autoencoder for mapping disease comorbidity in West Africa
A copula-integrated spatially regularized VAE is introduced to characterize geospatial comorbidity patterns of three childhood illnesses in West Africa from DHS data.
-
BeLink: Biomedical Entity Linking Meets Generative Re-Ranking
BeLink applies set-wise instruction-tuning to generative LLMs at the re-ranking stage of biomedical entity linking, reporting 3-24% accuracy gains and reduced inference time versus prior methods.
-
Cross-domain benchmarks reveal when coordinated AI agents improve scientific inference from partial evidence
Coordinated AI agents improve scientific inference from partial evidence in cross-domain tasks when single sources are incomplete, as demonstrated by AUROC gains in vector-borne disease and exoplanet benchmarks but tied performance in others.
-
Hardware-Software Co-Design of Scalable, Energy-Efficient Analog Recurrent Computations
BMRUs enable analog recurrent neural network hardware via discrete outputs that suppress noise 20-fold, with one-to-one parameter-to-circuit mapping and linear power scaling for recurrence.
-
CORTEG: Foundation Models Enable Cross-Modality Representation Transfer from Scalp to Intracranial Brain Recordings
Pretrained scalp-EEG foundation models can be transferred to ECoG via adapters and fine-tuning to match or exceed subject-specific baselines on regression tasks while requiring far less per-patient data.
-
A Computational Model of Message Sensation Value in Short Video Multimodal Features that Predicts Sensory and Behavioral Engagement
A computational MSV model built from short-video multimodal features positively predicts sensory engagement but shows an inverted-U relationship with behavioral engagement, validated on large unseen platform datasets.
-
BIAS: A Biologically Inspired Algorithm for Video Saliency Detection
BIAS is a biologically inspired video saliency model that integrates static and motion features via retina-like detection and multi-Gaussian fitting, outperforming baselines on DHF1K and anticipating traffic accidents up to 0.72 seconds early.
-
OpenPRC: A Unified Open-Source Framework for Physics-to-Task Evaluation in Physical Reservoir Computing
OpenPRC provides a schema-driven framework with five modules for GPU physics simulation, experimental vision ingestion, reservoir learning, information analysis, and physics-aware optimization to enable consistent PRC evaluation from simulations and real experiments.
-
From Human-Level AI Tales to AI Leveling Human Scales
Introduces a calibration framework for AI benchmarks using world-population probability levels on logarithmic scales derived from human test data and LLM extrapolation.
-
Adaptive digital twins for predictive decision-making: Online Bayesian learning of transition dynamics
An adaptive digital-twin framework learns state-transition probabilities online via Dirichlet-Multinomial Bayesian updates and recomputes finite-horizon maintenance policies, demonstrated on a simulated railway bridge.
-
Scalable Construction of Spiking Neural Networks using up to thousands of GPUs
A novel MPI-based construction method for spiking neural networks on multi-GPU clusters is introduced, with scaling demonstrated on two cortical models using point-to-point and collective communication.
-
RECAP: Transparent Inference-Time Emotion Alignment for Medical Dialogue Systems
RECAP is an inference-time framework using cognitive appraisal theory to enhance emotional alignment and transparency in medical dialogue systems across model scales.
-
An analysis of the effects of open science indicators on citations in the French Open Science Monitor
Preprints correlate with 19% higher citations, software sharing with 13.5%, data sharing with 14.3%, and open access with 8.6% in a citation prediction model applied to the French Open Science Monitor dataset.
-
Sketch of a novel approach to a neural model
The paper sketches a neuron-centric model of neuroplasticity that separates neural transmission from internal signal selection and storage within each neuron rather than relying solely on synaptic weights.
-
End-to-end plaque counting and virus titration from laboratory plate images with deep learning
SAM-derived models count plaques in virus titration plates with Pearson correlations of 0.92 and 0.88 to manual annotations on held-out data across viruses and plate formats.
-
A Game Theoretic Free Energy Analysis of Higher Order Synergy in Attention Heads of Large Language Models
Attention heads exhibit negative higher-order synergy (negative triple dividends), allowing pruning of redundant heads that cuts FLOPs by ~18% with only small perplexity increase.
-
Post-Moore Technologies for Plasma Simulation: A Community Roadmap
No single post-Moore technology replaces current HPC for plasma simulations, but FPGA-class accelerators offer near-term kernel offload, non-von Neumann architectures medium-term operator acceleration, and quantum computing long-term potential for warm dense matter microphysics.