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
DAG Structure Learning with Node-Varying Heavy-Tailed Disturbances Presents a DAG structure learning method that jointly estimates graph structure and node-varying heavy-tailed disturbance parameters. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on graphical model estimation, not AI deployment safety, hallucination, or jailbreaking.
DAG Structure Learning with Node-Varying Heavy-Tailed Disturbances Presents a DAG structure learning method that jointly estimates graph structure and node-varying heavy-tailed disturbance parameters. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Investigates causal DAG structure learning under heavy-tailed noise, not LLM benchmark evaluation.
DAG Structure Learning with Node-Varying Heavy-Tailed Disturbances Presents a DAG structure learning method that jointly estimates graph structure and node-varying heavy-tailed disturbance parameters. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on statistical graph learning algorithms, unrelated to small or efficient language models.
MORAD: Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design Introduces an online reinforcement learning framework optimizing multi-objective rewards across diverse target backbones for RNA inverse design. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on biological molecular generative modeling rather than efficient or small language models.
MORAD: Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design Introduces an online reinforcement learning framework optimizing multi-objective rewards across diverse target backbones for RNA inverse design. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Studies biological RNA sequence design, having no connection to AI safety or LLM guardrails.
MORAD: Reinforcement Learning with Multi-Objective Rewards for RNA Inverse Design Introduces an online reinforcement learning framework optimizing multi-objective rewards across diverse target backbones for RNA inverse design. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Deals with reinforcement learning for RNA sequence design, not LLM evaluation or benchmarks.
Structure-Aware Numerical Stabilization for Half-Precision Proximal Policy Optimization Proposes stabilization modules to enable half-precision FP16 proximal policy optimization training on commodity GPUs without numerical divergence. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Addresses RL numerical optimization details rather than AI alignment, safety, or deployment guardrails.
Structure-Aware Numerical Stabilization for Half-Precision Proximal Policy Optimization Proposes stabilization modules to enable half-precision FP16 proximal policy optimization training on commodity GPUs without numerical divergence. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on numerical stabilization for half-precision PPO reinforcement learning, not LLM benchmarking.
Sustainable Cloud-Fog-Edge Systems for Smart Buildings: Algorithmic Adaptation under Memory Constraints and Dynamic Orchestration Strategies Presents an edge-computing architecture using low-cost hardware and adaptive algorithms to enable autonomous data management in smart buildings. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on general edge IoT computing architectures rather than small language models or LLM compression.
Sustainable Cloud-Fog-Edge Systems for Smart Buildings: Algorithmic Adaptation under Memory Constraints and Dynamic Orchestration Strategies Presents an edge-computing architecture using low-cost hardware and adaptive algorithms to enable autonomous data management in smart buildings. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on IoT building orchestration, not AI safety, hallucination, or LLM guardrails.
Sustainable Cloud-Fog-Edge Systems for Smart Buildings: Algorithmic Adaptation under Memory Constraints and Dynamic Orchestration Strategies Presents an edge-computing architecture using low-cost hardware and adaptive algorithms to enable autonomous data management in smart buildings. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Covers IoT and smart building edge computing systems, not LLM evaluation methodology.
JEPA-kMMVAE: CSI Prediction for Time-Varying Wireless Channels Combines joint-embedding predictive architecture with a Markov variational autoencoder to predict wireless channel state information under noisy observations. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Deals with signal processing in wireless channels, not AI safety or deployment reliability.
JEPA-kMMVAE: CSI Prediction for Time-Varying Wireless Channels Combines joint-embedding predictive architecture with a Markov variational autoencoder to predict wireless channel state information under noisy observations. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on wireless channel state information prediction, completely unrelated to LLM evaluation.
JEPA-kMMVAE: CSI Prediction for Time-Varying Wireless Channels Combines joint-embedding predictive architecture with a Markov variational autoencoder to predict wireless channel state information under noisy observations. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Applies VAEs to telecommunications channel forecasting rather than efficient language model compression or serving.
SBNO : Schrödinger Bridge Neural Operator for the Forward and Inverse Problems under Missing Information Introduces a Schrödinger bridge diffusion matching method to solve forward and inverse partial differential equation problems under sparse observations. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on solving partial differential equations with diffusion bridges, not small language models or LLM compression.
SBNO : Schrödinger Bridge Neural Operator for the Forward and Inverse Problems under Missing Information Introduces a Schrödinger bridge diffusion matching method to solve forward and inverse partial differential equation problems under sparse observations. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Addresses scientific PDE modeling, not AI safety, alignment, or LLM deployment guardrails.
SBNO : Schrödinger Bridge Neural Operator for the Forward and Inverse Problems under Missing Information Introduces a Schrödinger bridge diffusion matching method to solve forward and inverse partial differential equation problems under sparse observations. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Studies surrogate neural operators for PDEs, not LLM benchmarks or evaluation methods.
DuoRate: Two-Rate Per-Expert Precision Allocation for Mixture-of-Experts Serving Allocates precision across MoE experts by separately handling resident storage footprint and active batch decode traffic via dynamic optimization. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Investigates inference efficiency and quantization, not AI safety, jailbreaks, or deployment reliability risks.
DuoRate: Two-Rate Per-Expert Precision Allocation for Mixture-of-Experts Serving Allocates precision across MoE experts by separately handling resident storage footprint and active batch decode traffic via dynamic optimization. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Focuses on precision allocation and serving efficiency for MoE, not LLM evaluation benchmarks.
PDE-PFN: Prior-Data Fitted Neural PDE Solver Proposes a transformer solver that uses in-context learning trained on approximate solutions to predict partial differential equation fields. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant In-context PDE solver framework, not LLM evaluation or benchmark design.
PDE-PFN: Prior-Data Fitted Neural PDE Solver Proposes a transformer solver that uses in-context learning trained on approximate solutions to predict partial differential equation fields. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on physics-informed in-context learning solvers, not small language model compression or efficiency.
PDE-PFN: Prior-Data Fitted Neural PDE Solver Proposes a transformer solver that uses in-context learning trained on approximate solutions to predict partial differential equation fields. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Presents a neural PDE solver, not AI safety, alignment, or deployment reliability controls.
AgentTune: A Unified Training Lifecycle for Tool-Using LLM Agents Presents an open-source library that unifies training, evaluation, reinforcement learning, and distillation workflows for tool-using language model agents. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on agent runtime, training, and workflows rather than small efficient model architectures.
AgentTune: A Unified Training Lifecycle for Tool-Using LLM Agents Presents an open-source library that unifies training, evaluation, reinforcement learning, and distillation workflows for tool-using language model agents. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Presents a broad framework for training and running agents, not an LLM benchmark or evaluation study.
Understanding Mamba-3 Through Oscillatory Dynamics Analyzes Mamba-3's input-dependent state rotations as oscillatory dynamics to explain their utility for tracking states and long-context language modeling. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Analyzes recurrent sequence model dynamics, not LLM evaluation benchmarks or methodologies.
Understanding Mamba-3 Through Oscillatory Dynamics Analyzes Mamba-3's input-dependent state rotations as oscillatory dynamics to explain their utility for tracking states and long-context language modeling. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Investigates recurrent oscillatory mechanics, not safety, alignment, or deployment guardrails.
Hidden Data Challenges in Clinical Machine Learning Identifies and evaluates temporal, semantic, and population preprocessing misalignments that cause data leakage and bias in clinical machine learning pipelines. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Audits clinical tabular preprocessing pipelines, not LLM evaluation or benchmark methodology.
Hidden Data Challenges in Clinical Machine Learning Identifies and evaluates temporal, semantic, and population preprocessing misalignments that cause data leakage and bias in clinical machine learning pipelines. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Studies clinical electronic health record preprocessing, not efficient models or inference techniques.
EmoGavel: Grounding-Aware Reinforcement Learning with Verifiable Rewards for Multimodal Emotion Preference Learning Proposes a reinforcement learning framework with verifiable rewards and error grounding to align multimodal emotion descriptions with human preferences. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on reinforcement learning for emotion preference alignment, not small efficient models.
EmoGavel: Grounding-Aware Reinforcement Learning with Verifiable Rewards for Multimodal Emotion Preference Learning Proposes a reinforcement learning framework with verifiable rewards and error grounding to align multimodal emotion descriptions with human preferences. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Method for training emotion preference models, not an LLM evaluation framework or benchmark study.
EmoGavel: Grounding-Aware Reinforcement Learning with Verifiable Rewards for Multimodal Emotion Preference Learning Proposes a reinforcement learning framework with verifiable rewards and error grounding to align multimodal emotion descriptions with human preferences. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on preference learning alignment for emotion tasks, not deployment guardrails or system safety.
Opening the Black Box of Classifier-Free Guidance via Information Bottleneck Analyzes classifier-free guidance in diffusion models through an information bottleneck framework to characterize when conditioning information is most effective. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Analyzes classifier-free guidance in diffusion, not LLM evaluation or benchmark validity.
Opening the Black Box of Classifier-Free Guidance via Information Bottleneck Analyzes classifier-free guidance in diffusion models through an information bottleneck framework to characterize when conditioning information is most effective. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Information-theoretic study of diffusion sampling schedules, not AI deployment safety or guardrails.
Opening the Black Box of Classifier-Free Guidance via Information Bottleneck Analyzes classifier-free guidance in diffusion models through an information bottleneck framework to characterize when conditioning information is most effective. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Studies diffusion sampling guidance schedules, not model compression, quantization, or small language models.
BLOKS: Scaling Kernel Solvers with Sketched Block Preconditioning Introduces a conjugate gradient solver for large-scale kernel ridge regression using sketched block preconditioning to improve memory efficiency and convergence. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Kernel regression optimization, not AI safety, alignment, or deployment guardrails.
BLOKS: Scaling Kernel Solvers with Sketched Block Preconditioning Introduces a conjugate gradient solver for large-scale kernel ridge regression using sketched block preconditioning to improve memory efficiency and convergence. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Scalable kernel solver algorithm, not LLM evaluation or benchmark reliability.
BLOKS: Scaling Kernel Solvers with Sketched Block Preconditioning Introduces a conjugate gradient solver for large-scale kernel ridge regression using sketched block preconditioning to improve memory efficiency and convergence. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Scales kernel ridge regression solvers, not deep model compression, quantization, or efficient LLM inference.
PlacePulse: Learning Transferable Place Representations from Structural Geographic Data Presents a geospatial foundation model that creates structural place embeddings from sociodemographic and business data across hexagonal spatial units. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Geospatial representation learning, not deployment safety, guardrails, or alignment risks.
PlacePulse: Learning Transferable Place Representations from Structural Geographic Data Presents a geospatial foundation model that creates structural place embeddings from sociodemographic and business data across hexagonal spatial units. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Introduces a geospatial foundation model and benchmark, not an evaluation method for language models.
PlacePulse: Learning Transferable Place Representations from Structural Geographic Data Presents a geospatial foundation model that creates structural place embeddings from sociodemographic and business data across hexagonal spatial units. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Geospatial embedding representation model, not small language model distillation, quantization, or efficient serving.
Model-free Bayesian Clustering and Multimodal Posterior Summarisation Develops a model-free Bayesian clustering method using martingale posteriors and Wasserstein summarization to bypass explicit likelihoods and MCMC sampling. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Bayesian clustering methodology, not LLM evaluation frameworks or benchmark metrics.
Model-free Bayesian Clustering and Multimodal Posterior Summarisation Develops a model-free Bayesian clustering method using martingale posteriors and Wasserstein summarization to bypass explicit likelihoods and MCMC sampling. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Statistical clustering algorithms, not AI safety, jailbreak defense, or deployment guardrails.
Model-free Bayesian Clustering and Multimodal Posterior Summarisation Develops a model-free Bayesian clustering method using martingale posteriors and Wasserstein summarization to bypass explicit likelihoods and MCMC sampling. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Addresses Bayesian clustering posterior computation, not model compression or efficient LLM deployment.
Why Linear Models Can Still Work in a Complex and Changing World Analyzes theoretically and empirically why linear models often out-predict complex nonlinear models under finite samples and environment distribution shifts. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Focuses on transfer risk of linear models, not operational LLM safety, red-teaming, or guardrails.
Why Linear Models Can Still Work in a Complex and Changing World Analyzes theoretically and empirically why linear models often out-predict complex nonlinear models under finite samples and environment distribution shifts. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Theoretical paper explaining linear model transferability, not LLM benchmark design or evaluation protocols.
Why Linear Models Can Still Work in a Complex and Changing World Analyzes theoretically and empirically why linear models often out-predict complex nonlinear models under finite samples and environment distribution shifts. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Analyzes linear versus nonlinear model generalization, not small language model compression or efficient serving.
Desertification Forecasting in Semi-Arid Regions: Using Deep Learning to Probe the Observability Limits of Land Degradation Dynamics Uses a convolutional LSTM on satellite data to evaluate the predictability and observational limits of year-to-year land degradation dynamics. Preprint	Computer vision for industrial applications via arxiv · 12 Oct, 16:29	Not relevant Focuses on satellite time-series land degradation forecasting, not industrial computer vision or inspection.
Desertification Forecasting in Semi-Arid Regions: Using Deep Learning to Probe the Observability Limits of Land Degradation Dynamics Uses a convolutional LSTM on satellite data to evaluate the predictability and observational limits of year-to-year land degradation dynamics. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Satellite land degradation forecasting study, not general LLM evaluation or benchmark design.
Desertification Forecasting in Semi-Arid Regions: Using Deep Learning to Probe the Observability Limits of Land Degradation Dynamics Uses a convolutional LSTM on satellite data to evaluate the predictability and observational limits of year-to-year land degradation dynamics. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Application of ConvLSTM to Earth observation data, not model compression or efficient serving.
Desertification Forecasting in Semi-Arid Regions: Using Deep Learning to Probe the Observability Limits of Land Degradation Dynamics Uses a convolutional LSTM on satellite data to evaluate the predictability and observational limits of year-to-year land degradation dynamics. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Environmental forecasting modeling, not AI safety, alignment, guardrails, or deployment reliability risks.
Accelerating Linear Bandits with Persistent Candidate Sets Introduces an efficient linear bandit algorithm that maintains a small candidate set of actions to reduce per-round evaluation costs. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Bandit regret minimization analysis, not AI safety, hallucination mitigation, or deployment guardrails.
Accelerating Linear Bandits with Persistent Candidate Sets Introduces an efficient linear bandit algorithm that maintains a small candidate set of actions to reduce per-round evaluation costs. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Theoretical bandit algorithm acceleration, not LLM evaluation or benchmarking frameworks.
Accelerating Linear Bandits with Persistent Candidate Sets Introduces an efficient linear bandit algorithm that maintains a small candidate set of actions to reduce per-round evaluation costs. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on linear multi-armed bandit optimization, not small language models or deep model efficiency.
One Thing Leads to Another: The Essentials of Time Series Causal Discovery Provides a comprehensive tutorial on time series causal discovery methods, covering underlying principles, evaluation techniques, and practical applications. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Overview of time-series causality principles, not model compression or efficient LLM inference.
One Thing Leads to Another: The Essentials of Time Series Causal Discovery Provides a comprehensive tutorial on time series causal discovery methods, covering underlying principles, evaluation techniques, and practical applications. Preprint	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Tutorial on time series causal discovery methods, not LLM evaluation benchmarks or testing protocols.
One Thing Leads to Another: The Essentials of Time Series Causal Discovery Provides a comprehensive tutorial on time series causal discovery methods, covering underlying principles, evaluation techniques, and practical applications. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Tutorial on time-series causal inference methods, not deployment safety, guardrails, or red-teaming.
PPG-Bench: A Unified Benchmark for PPG Foundation Models Introduces a unified benchmark evaluating five public photoplethysmography foundation models across nine datasets and 26 clinical and wearable tasks. Preprint	Small and efficient models via arxiv · 12 Oct, 16:29	Not relevant Focuses on benchmarking photoplethysmography foundation models, not small efficient model techniques.
PPG-Bench: A Unified Benchmark for PPG Foundation Models Introduces a unified benchmark evaluating five public photoplethysmography foundation models across nine datasets and 26 clinical and wearable tasks. Preprint	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Benchmarking representation quality on biosignals, not deployment safety, guardrails, or alignment.
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	LLM evaluation and benchmarks via arxiv · 12 Oct, 16:29	Not relevant Statistical learning theory on risk bounds, not empirical LLM evaluation or benchmark frameworks.
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	AI safety and reliability in deployment via arxiv · 12 Oct, 16:29	Not relevant Pure statistical learning theory, not deployed AI safety, guardrails, or hallucination mitigation.

15,121–15,180 of 17,313← Previous Next →