This study explores the application of Masked Autoencoder (MAE) pretraining for downhole drilling metric prediction, addressing the data asymmetry in drilling telemetry. Using real well drilling data, MAE reduced the test mean absolute error by 19.8% relative to supervised GRU baselines for Total Mud Volume prediction.
AdaGraph is a graph-native clustering algorithm from the Structure-Centric Machine Learning (SC-ML) paradigm, which fundamentally dissolves the curse of dimensionality by replacing geometry-centric computation with topology-based computation. Operating within kNN graph topology, it requires no a priori specification of cluster numbers, handles noise, and scales effectively.
This study develops a hybrid predictive framework using machine learning (CatBoost, SHAP) and logistic regression to identify and quantify risk factors contributing to injury severity in tree-involved traffic crashes. It analyzes CRSS data from 2020-2023 to understand high-energy impacts often resulting in fatal or severe injuries.
This paper addresses constrained variational inequality problems with functional constraints, proposing mirror descent-type algorithms. These algorithms are analyzed for their optimal convergence rate for problems with bounded and monotone operators and Lipschitz convex functional constraints.
This paper introduces a residual gap-aware transformer for forecasting 24-month Alzheimer's disease progression using ADNI clinical and biomarker histories. The research analyzes changes in CDR-SB scores, anchoring samples at mild cognitive impairment visits.
Contrastive learning aims to preserve relational structure in sample representations by reflecting a similarity graph. This paper interprets weighted InfoNCE objectives as Distance Geometry Problems, providing a unified geometric framework and exact characterizations of optimal embeddings, revealing how class imbalance affects inter-class similarities in SupCon.
This research analyzes three KV cache quantization schemes (KV, KQV, QKQV) and their impact on inner product variance, especially how QJL on K inflates it, amplified by softmax. Empirical findings highlight KQV's superior performance at a budget of n=4, an unconditional K-V asymmetry where QKQV is consistently worse than KQV in KL divergence, and budget-dependent crossovers for geometric K reconstruction.
This work introduces a classification of learning dynamics (scalar-reducible vs. scalar-irreducible) and demonstrates that scalar-irreducible dynamics enable endogenous regime switching, crucial for autonomous intelligence. It proposes a new dynamical paradigm for regime exploration without external scheduling.
This paper argues that for collective decision-making based on free-form text, embeddings should measure "preferential similarity" rather than "semantic similarity". Existing embeddings capture a coarse preference signal, but fail when this correlation breaks, a problem formalized as an invariance issue.
This research introduces a scalable computational approach to measure manner and result verbs, a crucial distinction for developmental language studies. It leverages large language models for sentence annotations and trains a RoBERTa-based classifier, demonstrating promising performance on evaluation datasets.
MemQ integrates TD($\lambda$) eligibility traces with memory Q-values, propagating credit backward through a provenance DAG to account for memory dependencies. This approach significantly improves LLM agents' ability to accumulate and retrieve experience, achieving high success rates across various benchmarks.
A hierarchical ensemble pipeline is proposed for anomaly detection in multivariate telemetry data from the European Space Agency (ESA). This method, integrating various feature extraction and modeling techniques, shows strong generalization and effectiveness in detecting subtle anomalies in satellite telemetry.
The paper introduces ATOM, a multi-agent framework for multi-objective molecular optimization employing a tree-structured search. Agents coordinate along different paths of the tree to maintain and compare alternative molecular evolution trajectories, supported by a global memory.
This paper proposes a "lookahead drifting model" for distribution mapping, which enhances image generation performance via one-step neural functional evaluation. The model computes a set of drifting terms sequentially at each training iteration, utilizing positive samples and model outputs to capture higher-order gradient information.
This paper introduces Foundation Preserving LoRA (FoLoRA), an optimization framework that addresses the degradation of nontarget capabilities during finetuning of foundation models. It uses a generalized Rayleigh quotient to balance task utility and forgetting penalty, guiding updates to preserve pretraining knowledge.
This paper explores training language models to forecast the empirical success of research ideas by evaluating pairs of ideas against objective outcomes. SFT significantly boosts performance beyond GPT-5, and RLVR can train models to discover interpretable reasoning paths for this forecasting task.
We propose PathBoost, a gradient tree boosting method for graph-level classification and regression, which learns discriminative path-based features directly from the input graph structure. This method introduces adaptations for binary classification, incorporates multiple node and edge attributes, and automatically selects anchor nodes, outperforming or matching graph neural networks and graph kernel approaches on several benchmark datasets.
This paper introduces Deployment-Time Learning (DTL) as a new stage for LLMs, allowing them to continually adapt from experience post-training without modifying core parameters. It presents CASCADE, a framework that uses an explicit, evolving episodic memory for LLM agents, formalizing experience reuse as a contextual bandit problem with no-regret guarantees.
BaLoRA is a Bayesian extension of LoRA that enhances the accuracy of large-scale model adaptation. This novel approach not only quantifies uncertainty but also significantly narrows the performance gap with full fine-tuning.
This research paper introduces a coin-flip model to predict the performance of symbolic and prompt-based LLM programs using a few in-domain examples and a performance prior. It finds that symbolic programs exhibit an "all or nothing" performance prior, while prompt programs have a diffuse prior.