Every paper checked — nothing hidden

All papers

Every paper checked for brief #4 of 12–19 Oct 2026 — nothing is hidden. Brief papers first.

All17,313 In brief24 More (on-topic)3,333 Not relevant13,754 Not assessed87 Older352 Could not confirm93 Blogs with no paper266 All blogs284

Paper	Topic	Relevance
Learning by Compression: Risk Estimates Attain the Law-of-Large-Numbers Rate Proves that risk estimates derived from preferent compression schemes achieve standard law-of-large-numbers rates without excess logarithmic factors. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Theoretical sample-compression risk bounds, not practical model compression, quantization, or efficient inference.
Harmonic analysis and circuit approximation of Sobolev functions on quantum input spaces Derives theoretical approximation rates for Sobolev functions on quantum state and gate spaces using harmonic analysis and finite-depth quantum circuits. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on quantum Sobolev function approximation, not LLM benchmarks or evaluation trustworthiness.
Harmonic analysis and circuit approximation of Sobolev functions on quantum input spaces Derives theoretical approximation rates for Sobolev functions on quantum state and gate spaces using harmonic analysis and finite-depth quantum circuits. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Theoretical quantum circuit approximation rates, not classical small language models or efficient inference.
Harmonic analysis and circuit approximation of Sobolev functions on quantum input spaces Derives theoretical approximation rates for Sobolev functions on quantum state and gate spaces using harmonic analysis and finite-depth quantum circuits. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Addresses quantum circuit mathematical approximation theory, not real-world AI deployment safety or robustness.
A Scalable Learned Iterative Solver for Nonlinear MPC via Temporal Decomposition Introduces a learned primal-dual solver for nonlinear model predictive control that uses temporal decomposition to scale independently of the prediction horizon. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Concerns optimal control solver architectures, not efficient small language models or LLM inference serving.
A Scalable Learned Iterative Solver for Nonlinear MPC via Temporal Decomposition Introduces a learned primal-dual solver for nonlinear model predictive control that uses temporal decomposition to scale independently of the prediction horizon. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on numerical control solvers, not AI deployment safety, alignment, or guardrail systems.
A Scalable Learned Iterative Solver for Nonlinear MPC via Temporal Decomposition Introduces a learned primal-dual solver for nonlinear model predictive control that uses temporal decomposition to scale independently of the prediction horizon. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Addresses learned optimization for nonlinear predictive control, not LLM benchmarks or evaluation methods.
FedPCD: Priority-Constrained Federated Aggregation Proposes a federated optimization framework that preserves global aggregate objectives while applying minimal corrections to meet individual client performance constraints. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Covers federated learning optimization across clients, not model compression, quantization, or efficient serving.
FedPCD: Priority-Constrained Federated Aggregation Proposes a federated optimization framework that preserves global aggregate objectives while applying minimal corrections to meet individual client performance constraints. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on federated optimization algorithms, not LLM evaluation methods or benchmark datasets.
FedPCD: Priority-Constrained Federated Aggregation Proposes a federated optimization framework that preserves global aggregate objectives while applying minimal corrections to meet individual client performance constraints. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Proposes a federated aggregation constraint algorithm, not LLM safety, guardrails, or deployment vulnerability mitigation.
Asymptotic Analysis of Discrete-Time Self-Attention Dynamics in Transformers Analyzes the discrete-time dynamical system of transformer self-attention, classifying its equilibria and showing consensus alignment with the value matrix's principal eigenvector. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Studies mathematical dynamical stability of attention layers, not deployment safety, alignment, or guardrails.
Asymptotic Analysis of Discrete-Time Self-Attention Dynamics in Transformers Analyzes the discrete-time dynamical system of transformer self-attention, classifying its equilibria and showing consensus alignment with the value matrix's principal eigenvector. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Investigates theoretical equilibria of attention dynamics, not efficient inference or compact model architectures.
Asymptotic Analysis of Discrete-Time Self-Attention Dynamics in Transformers Analyzes the discrete-time dynamical system of transformer self-attention, classifying its equilibria and showing consensus alignment with the value matrix's principal eigenvector. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Theoretical analysis of self-attention dynamical systems, not LLM evaluation benchmarks or methodologies.
Simulating Training Spikes of Deep Neural Networks by Layer-wise Random Feature Models Simulates transformer training loss spikes using quadratic models with heavy-tailed inputs and derives an adaptive optimizer hyperparameter schedule to prevent them. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on training stability dynamics, not model compression, distillation, quantization, or efficient inference.
Simulating Training Spikes of Deep Neural Networks by Layer-wise Random Feature Models Simulates transformer training loss spikes using quadratic models with heavy-tailed inputs and derives an adaptive optimizer hyperparameter schedule to prevent them. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Studies pre-training loss spike dynamics theoretically, not LLM benchmark design or evaluation trustworthiness.
Simulating Training Spikes of Deep Neural Networks by Layer-wise Random Feature Models Simulates transformer training loss spikes using quadratic models with heavy-tailed inputs and derives an adaptive optimizer hyperparameter schedule to prevent them. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Examines training loss spikes, not deployment safety, jailbreaks, prompt injection, or operational reliability.
Probabilistic Ship Trajectory Forecasting using Transformer Models on AIS Data Evaluates Transformer models trained with quantile loss on historical vessel position data to produce probabilistic ship trajectory forecasts. Preprint	Computer vision for industrial applications via arxiv · 12 Oct, 16:29	Not relevant Uses AIS coordinates and positional heat maps, not industrial computer vision or defect inspection.
Probabilistic Ship Trajectory Forecasting using Transformer Models on AIS Data Evaluates Transformer models trained with quantile loss on historical vessel position data to produce probabilistic ship trajectory forecasts. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Studies vessel collision avoidance forecasting, not LLM deployment safety, guardrails, or alignment.
Probabilistic Ship Trajectory Forecasting using Transformer Models on AIS Data Evaluates Transformer models trained with quantile loss on historical vessel position data to produce probabilistic ship trajectory forecasts. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on trajectory prediction models, not efficient model compression or serving.
Probabilistic Ship Trajectory Forecasting using Transformer Models on AIS Data Evaluates Transformer models trained with quantile loss on historical vessel position data to produce probabilistic ship trajectory forecasts. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Evaluates maritime vessel trajectory forecasting, not LLM evaluation or benchmarks.
Rationally Linearized MLPs Linearizes multilayer perceptrons using low-degree rational approximations of activation functions to reduce inference costs across 3D models and language models. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Deals with architectural inference efficiency, not AI safety, jailbreaks, or guardrails.
Rationally Linearized MLPs Linearizes multilayer perceptrons using low-degree rational approximations of activation functions to reduce inference costs across 3D models and language models. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on MLP linearization techniques, not benchmark creation or LLM evaluation.
Simultaneous Competitive and Regret Guarantees for TCP Acknowledgment Presents an online randomized TCP acknowledgment algorithm that provides simultaneous competitive ratio bounds and sublinear regret against the best expert. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Addresses theoretical online algorithms for TCP, not small or efficient models.
Simultaneous Competitive and Regret Guarantees for TCP Acknowledgment Presents an online randomized TCP acknowledgment algorithm that provides simultaneous competitive ratio bounds and sublinear regret against the best expert. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Online algorithms paper for networking TCP ACK, unrelated to LLM benchmarks.
Simultaneous Competitive and Regret Guarantees for TCP Acknowledgment Presents an online randomized TCP acknowledgment algorithm that provides simultaneous competitive ratio bounds and sublinear regret against the best expert. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Theoretical competitive analysis for network packets, unrelated to AI safety.
$L_0$ Feature Selection via Monte Carlo Tree Search in High-Dimensional Financial Regression Applies Monte Carlo Tree Search with generalized rollout policy adaptation to optimize out-of-sample feature selection for high-dimensional financial regression. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Deals with financial feature selection via search, not LLM benchmarks or evaluation.
$L_0$ Feature Selection via Monte Carlo Tree Search in High-Dimensional Financial Regression Applies Monte Carlo Tree Search with generalized rollout policy adaptation to optimize out-of-sample feature selection for high-dimensional financial regression. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Applies tree search to tabular regression, unrelated to AI deployment safety.
$L_0$ Feature Selection via Monte Carlo Tree Search in High-Dimensional Financial Regression Applies Monte Carlo Tree Search with generalized rollout policy adaptation to optimize out-of-sample feature selection for high-dimensional financial regression. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on sparse regression feature selection, not efficient model compression or serving.
Probabilistic Forecasting of Electric Vehicle Charging in a Time-varying Hierarchical Infrastructure Combines hierarchical forecasting with conformal prediction under missing data to forecast electric vehicle charging demand with valid prediction intervals across evolving networks. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Studies conformal prediction for EV demand forecasting, not LLM evaluation or benchmarks.
Probabilistic Forecasting of Electric Vehicle Charging in a Time-varying Hierarchical Infrastructure Combines hierarchical forecasting with conformal prediction under missing data to forecast electric vehicle charging demand with valid prediction intervals across evolving networks. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Covers infrastructure forecasting uncertainty, not AI deployment safety or guardrails.
Probabilistic Forecasting of Electric Vehicle Charging in a Time-varying Hierarchical Infrastructure Combines hierarchical forecasting with conformal prediction under missing data to forecast electric vehicle charging demand with valid prediction intervals across evolving networks. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on time-series conformal forecasting, not small models or efficient inference.
Lower Bounds for Randomized Linear-Oracle Optimization over Strongly Convex Sets Investigates theoretical lower bounds for randomized linear minimization oracle optimization algorithms over strongly convex sets. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Pure mathematical optimization theory, unrelated to AI safety or reliability.
Lower Bounds for Randomized Linear-Oracle Optimization over Strongly Convex Sets Investigates theoretical lower bounds for randomized linear minimization oracle optimization algorithms over strongly convex sets. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on mathematical optimization lower bounds, not small models or efficient inference.
Lower Bounds for Randomized Linear-Oracle Optimization over Strongly Convex Sets Investigates theoretical lower bounds for randomized linear minimization oracle optimization algorithms over strongly convex sets. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Theoretical optimization lower bounds paper, unrelated to LLM evaluation.
EchoTab: Learning Measurement-Grounded Representations for Echocardiographic Diagnosis through Tabular Reconstruction Proposes a measurement-grounded representation framework using tabular tokens to link multi-view echocardiograms to clinical diagnostic tasks. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Proposes medical imaging representation learning, not LLM benchmarks or evaluation.
EchoTab: Learning Measurement-Grounded Representations for Echocardiographic Diagnosis through Tabular Reconstruction Proposes a measurement-grounded representation framework using tabular tokens to link multi-view echocardiograms to clinical diagnostic tasks. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on echocardiographic representation learning, not efficient model inference or compression.
EchoTab: Learning Measurement-Grounded Representations for Echocardiographic Diagnosis through Tabular Reconstruction Proposes a measurement-grounded representation framework using tabular tokens to link multi-view echocardiograms to clinical diagnostic tasks. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Deals with medical imaging diagnostic representations, not AI safety or deployment guardrails.
A small cognitive core explains human behavior at scale Presents a cognitive architecture based on episodic memory hierarchies and context-dependent error updates to predict human choice behavior. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Investigates human choice psychology, not AI deployment reliability or guardrails.
A small cognitive core explains human behavior at scale Presents a cognitive architecture based on episodic memory hierarchies and context-dependent error updates to predict human choice behavior. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Presents a cognitive architecture, not methods for compressing or serving efficient models.
A small cognitive core explains human behavior at scale Presents a cognitive architecture based on episodic memory hierarchies and context-dependent error updates to predict human choice behavior. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Cognitive psychology computational model of human choices, not an LLM evaluation benchmark.
CorrFlow: Exact and Permutation-Equivariant Flow Matching on the Correlation Elliptope Proposes an unconstrained parameterization and equivariant flow matching model for generating valid correlation matrices. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on financial matrix generation, not AI deployment safety or guardrails.
CorrFlow: Exact and Permutation-Equivariant Flow Matching on the Correlation Elliptope Proposes an unconstrained parameterization and equivariant flow matching model for generating valid correlation matrices. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Deals with geometry of correlation matrix generation, not small or efficient models.
CorrFlow: Exact and Permutation-Equivariant Flow Matching on the Correlation Elliptope Proposes an unconstrained parameterization and equivariant flow matching model for generating valid correlation matrices. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Introduces a flow matching method for correlation matrices, not an LLM benchmark.
Vocal Intensity Conversion with Continuous SPL Control via Adversarial Training Presents a cycle-consistent adversarial model enabling continuous sound pressure level adjustments for speech vocal intensity conversion. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Covers speech synthesis and vocal intensity conversion, not LLM evaluation or benchmarks.
Vocal Intensity Conversion with Continuous SPL Control via Adversarial Training Presents a cycle-consistent adversarial model enabling continuous sound pressure level adjustments for speech vocal intensity conversion. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Deals with audio signal intensity conversion, not AI deployment safety or reliability.
Vocal Intensity Conversion with Continuous SPL Control via Adversarial Training Presents a cycle-consistent adversarial model enabling continuous sound pressure level adjustments for speech vocal intensity conversion. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on audio GAN conversion techniques, not model compression or efficient inference.
Some remarks on the connection between generative diffusion and optimal transport Analyzes theoretical connections between generative diffusion models, flow matching, and entropic optimal transport formulations. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Theoretical analysis of diffusion and optimal transport, not LLM evaluation or benchmarks.
Some remarks on the connection between generative diffusion and optimal transport Analyzes theoretical connections between generative diffusion models, flow matching, and entropic optimal transport formulations. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Mathematical study of generative diffusion dynamics, not efficient inference or small models.
Some remarks on the connection between generative diffusion and optimal transport Analyzes theoretical connections between generative diffusion models, flow matching, and entropic optimal transport formulations. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on optimal transport theory, not practical AI safety or reliability in deployment.
Kernel-based External Validity of Experimental Studies Develops a kernel-based maximum mean discrepancy framework to formalize and estimate the external validity and replicability of experimental studies. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Theoretical statistics framework for experimental validity, unrelated to efficient model inference.
Kernel-based External Validity of Experimental Studies Develops a kernel-based maximum mean discrepancy framework to formalize and estimate the external validity and replicability of experimental studies. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Addresses theoretical statistical external validity using kernels, not LLM evaluation benchmarks.
Kernel-based External Validity of Experimental Studies Develops a kernel-based maximum mean discrepancy framework to formalize and estimate the external validity and replicability of experimental studies. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on statistical replication theory rather than AI deployment safety, alignment, or guardrails.
JitterVLA: A Replay Benchmark for Motion Quality in VLA-Driven Human-Robot Manipulation Introduces a replay benchmark to evaluate robot motion smoothness and controller latency in vision-language-action manipulation pipelines. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on motion smoothing in robot controllers, not model compression or efficient inference.
JitterVLA: A Replay Benchmark for Motion Quality in VLA-Driven Human-Robot Manipulation Introduces a replay benchmark to evaluate robot motion smoothness and controller latency in vision-language-action manipulation pipelines. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Benchmark focuses on robotic motion quality in controllers, not LLM evaluation methods.
JitterVLA: A Replay Benchmark for Motion Quality in VLA-Driven Human-Robot Manipulation Introduces a replay benchmark to evaluate robot motion smoothness and controller latency in vision-language-action manipulation pipelines. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Studies physical robot trajectory controller delays rather than software AI safety or guardrails.
JitterVLA: A Replay Benchmark for Motion Quality in VLA-Driven Human-Robot Manipulation Introduces a replay benchmark to evaluate robot motion smoothness and controller latency in vision-language-action manipulation pipelines. Preprint	Computer vision for industrial applications via arxiv · 12 Oct, 16:29	Not relevant Studies trajectory jitter and controller smoothing in robotics, not industrial vision or inspection.
Private and Distributionally Robust Federated Multiclass Classification Develops a differentially private and distributionally robust ADMM framework for multiclass classification across distributed client data. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Theoretical federated learning privacy framework, not deployment safety, guardrails, or LLM hallucination.
Private and Distributionally Robust Federated Multiclass Classification Develops a differentially private and distributionally robust ADMM framework for multiclass classification across distributed client data. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Addresses federated optimization with differential privacy, not small language models or efficient inference.
Private and Distributionally Robust Federated Multiclass Classification Develops a differentially private and distributionally robust ADMM framework for multiclass classification across distributed client data. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on federated classification algorithms, not LLM evaluation or benchmarks.
PANDA: Prior-guided Attentional Dual-path Architecture Presents a dual-path architecture combining normalized attention and incidence priors for hypergraph message-passing neural networks. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on hypergraph attention mechanisms, not efficient language model compression or inference.

15,181–15,240 of 17,313← Previous Next →