Research
On-device research index

arXiv research

A locally-built, LLM-digested index of recent arXiv papers in quant finance, geometry/topology, and statistical ML — keyword search served straight from SQLite on this machine.

169,341 papers · 148 categories

Trend · papers per month

212423635846 · Jun 202019922001200920182026
48 results for writing improvement

Analyzed writing style changes in Danish high school students.

problem Detecting global development trends and identifying at-risk students in high school writing.
method Used a Siamese neural network to compute essay similarity and clustered student profiles.
result High school students' writing styles become less similar as they progress, with some students showing significant improvement and others limited development or setbacks.

Firms delay write-downs for adverse macroeconomic and industry outcomes but not for firm-specific issues.

problem Timeliness of write-downs for adverse macroeconomic and industry outcomes versus firm-specific issues.
method Comparative analysis of write-downs driven by macroeconomic and industry outcomes versus firm-specific outcomes.
result Firms delay write-downs for adverse macroeconomic and industry outcomes but not for firm-specific issues.

SAGE improves memory efficiency by selectively adding, merging, or ignoring new facts.

problem Efficiently managing new facts in agentic LLMs to avoid costly write-time reasoning.
method SAGE uses a von Mises-Fisher-based density estimator to score and route candidate facts.
result SAGE achieves the best average token-F1 on LoCoMo and reduces add-phase API cost by 3.4x on GPT-4o-mini.

Identifies beneficial interventions from observational data, even in misspecified settings.

problem Identifying beneficial interventions from observational data, especially in applications where harmful interventions are worse.
method Develops efficient algorithms to identify optimal intervention policies from limited data, assuming covariates can be precisely adjusted.
result Empirically, our approach identifies good interventions in gene perturbation and writing improvement applications.

Deep learning improves handwriting style transfer and extraction.

problem Improving handwriting style transfer and extraction using deep neural networks.
method Used a deep conditioned autoencoder on IRON-OFF handwriting data-set to explore style transfer and extraction.
result Improved metrics of state-of-the-art methods by a large margin in style transfer and extraction experiments.

Survival analysis models predict loan write-off risk under IFRS 9.

problem Estimating loan write-off probabilities in credit risk modeling.
method Discrete-time hazard model and conditional inference survival tree compared to cross-sectional logistic regression.
result Discrete-time hazard model outperforms other two-stage LGD-models.

Topic model captures health journeys of multiple authors.

problem Challenges in topic modeling health journals due to asynchronous writing.
method Dynamic Author-Persona topic model (DAP) with regularized variational inference.
result Significant improvements over competing models, especially with regularization.

Improves writer identification with unlabeled data and weighted label smoothing.

problem Offline writer identification requires labeled data, which is costly and time-consuming.
method Proposed a semi-supervised feature learning pipeline with weighted label smoothing regularization.
result Significantly improved baseline performance on writer identification datasets.

Low-rank training improves neural network training on edge devices with non-volatile memory.

problem Training neural networks on edge devices with non-volatile memory, especially in terms of write density and auxiliary memory.
method Low-rank training scheme to address write density and auxiliary memory limitations.
result The low-rank training technique outperforms standard SGD in accuracy and weight writes.

New method optimizes memory usage in neural networks, improving sequential learning.

problem Current memory models in neural networks waste memory and computation.
method Formulated an optimization problem to maximize information storage, introduced Cached Uniform Writing.
result Proved Cached Uniform Writing optimizes memory usage and outperforms other methods.

HawkesLLM models text generation with temporal influence, improving semantic alignment under limited memory.

problem Path-dependent uncertainty in agentic text-simulation systems.
method HawkesLLM framework separates temporal influence modeling from text generation, using a multivariate Hawkes process and a language model.
result HawkesLLM improves late-stage semantic alignment under a compact prompt-memory budget.

AutoGraph improves Python coding for machine learning, combining ease and performance.

problem The trade-off between ease of writing and scalability in machine learning code.
method Source code transformation using staged programming in Python, delaying type-dependent decisions until runtime.
result Usability improvements with no performance loss compared to native TensorFlow graphs.

A new presentation of the nn-string braid group BnB_n is studied. Using it, a new solution to the word problem in BnB_n is obtained which retains most of the desirable features of the Garside-Thurston solution, and at the same time makes possible certain computational improvements. We also give a related solution to t…

1997-12-02abs ↗pdf ↗

A wrist-worn device system for user authentication using writing behavior analysis.

problem User authentication through writing behavior for wearable devices.
method Dynamic Time Wrapping and Savitzky-Golay filter for fine-grained writing metrics.
result The proposed system achieves high accuracy in user identification with low false-positive and false-negative rates.

A new framework for efficient sequence maps using Bayesian filtering and covariance.

problem Designing efficient recurrent sequence maps from explicit memory assumptions.
method Design-model framework, exact Bayesian filtering, query-dependent readout, linear-Gaussian instantiation.
result Improved robustness and retrieval performance across various benchmarks.

Paper proposes SA-VAE for generating stylized Chinese characters.

problem Automatic generation of stylized Chinese characters is challenging.
method Proposes Style-Aware Variational Auto-Encoder (SA-VAE) to capture content and style components.
result Shows powerful one-shot/low-shot generalization ability.

Study evaluates three position sizing methods for put-writing on S&P 500 Index options.

problem Underdeveloped practical implementation of short-dated volatility-selling strategies.
method Kelly criterion, VIX-based volatility scaling, hybrid method.
result Ultra-short-dated, out-of-the-money options deliver superior risk-adjusted returns.

The study predicts how discussions in mental disorder Reddit communities affect users' emotional states.

problem Improving mental health conditions through social support analysis.
method Text embedding techniques and RNNs for predicting emotional tone shifts.
result Users' emotional states can improve due to social support, as evidenced by positive comments following negative posts.

AI agents improve forecast combination in empirical economics.

problem Hidden researcher degrees of freedom in AI-generated code.
method Adapted agent-loop architecture to empirical economics, added holdout evaluation.
result Independent agent searches find better forecast methods than benchmarks.

The study examines how knots relate based on a specific map property.

problem Understanding the complexity of knots through a map property.
method Examines knots through a degree 1 map property between their exteriors.
result Supports the idea that more complex knots have a higher map property.

Dual neural network improves treatment recommendations from medical history.

problem Improving treatment recommendations from patient medical history.
method Memory-augmented neural network with dual controllers.
result Dual controller write-protected memory-augmented neural network outperformed traditional methods.

The paper provides bounds for the ropelength of a link in terms of the crossing numbers of its split components. As in earlier papers, the bounds grow with the square of the crossing number; however, the constant involved is a substantial improvement on previous results. The proof depends essentially on writing links i…

2002-10-16abs ↗pdf ↗

Improves sampling from complex hierarchical models using HMC and automatic marginalization.

problem Sampling from complex hierarchical models is difficult for HMC.
method Proposes automatic marginalization as part of the sampling process using HMC in a graphical model extracted from a PPL.
result Significantly improves sampling from real-world hierarchical models.

Efficiently processes dynamic inputs in AI writing assistants with incremental computation.

problem Efficiently updating AI models in real-time with dynamic inputs.
method Incremental computing using vector quantization to filter and reuse intermediate values in neural networks.
result Comparable accuracy with 12.1X fewer operations for processing dynamic inputs.

CLARA generates clinical reports from raw inputs, improving accuracy and efficiency.

problem Generating accurate and detailed clinical reports from raw inputs is time-consuming and error-prone.
method Interactive method that generates reports sentence by sentence based on doctors' anchor words and partially completed sentences.
result CLARA achieves significant improvements in report generation accuracy and efficiency.

DPBD simplifies labeling functions through interactive demonstrations.

problem Difficulty in writing labeling functions for large-scale labeled training data.
method Data Programming by Demonstration (DPBD) framework using interactive demonstrations.
result Ruler system generates labeling rules more easily and with higher user satisfaction.

Study finds AI-generated financial advice influences life cycle investing patterns.

problem Understanding how AI-generated financial advice impacts life cycle investing.
method Sentiment analysis of prompts from AI-generated financial advice and simulation of lifetime effects.
result AI-generated financial advice leads to life cycle investing patterns, influenced by gender and AI experience.

In this paper, a generalized version of Morton's formula is proved. Using this formula, one can write down the colored Jones polynomials of cabling of an knot in terms of the colored Jones polynomials of the original knot.

2008-10-09abs ↗pdf ↗

The starting point of our analysis is an old idea of writing an eigenfunction expansion for a heat kernel considered in the case of a hypoelliptic heat kernel on a nilpotent Lie group GG. One of the ingredients of this approach is the generalized Fourier transform. The formula one gets using this approach is explicit …

2015-05-15abs ↗pdf ↗

This part 2 discusses virtual fundamental chain and cycle technique for K-systems.

problem Foundation of virtual fundamental chain and cycle technique for K-systems.
method Consider a system of spaces with Kuranishi structures and their simultaneous perturbations.
result Discuss the virtual fundamental chain and cycle technique for K-systems.