Context2Name predicts natural names from minified code, improving code readability.
problem Minified code makes it hard to understand natural names in JavaScript.
method Combines static analysis and neural networks to predict natural names.
result Successfully predicts 47.5% of all minified identifiers in real-world code.
The simulator is an R package that streamlines the process of performing simulations by creating a common infrastructure that can be easily used and reused across projects. Methodological statisticians routinely write simulations to compare their methods to preexisting ones. While developing ideas, there is a temptatio…
Automatic differentiation (AD) is an essential primitive for machine learning programming systems. Tangent is a new library that performs AD using source code transformation (SCT) in Python. It takes numeric functions written in a syntactic subset of Python and NumPy as input, and generates new Python functions which c…
Transformer model improves source code summarization.
problem Generating readable summaries of source code.
method Transformer model with self-attention mechanism for code representation.
result Transformer model outperforms state-of-the-art techniques.
SPX optimizes multiple graph drawing metrics for better readability.
problem Graph drawing algorithms often optimize one metric at a time, leading to suboptimal layouts.
method Introduces Stress-Plus-X (SPX) framework that optimizes stress, crossings, angles, and upwardness simultaneously.
result SPX achieves results close to state-of-the-art algorithms that optimize metrics individually.
RL for image captions improved with a language prior.
problem Learning biases and large sample space issues in RL image captioning.
method Added a language prior to constrain the action space.
result RL with the language prior module performs better in readability and speed.
SigD2 reduces noisy rules in rule-based classifiers for better accuracy and readability.
problem Redundant and noisy rules in rule-based classifiers reduce model accuracy and readability.
method Two-stage pruning strategy and ensemble methods (bagging and boosting) to reduce noise and improve model performance.
result SigD2 and ACboost ensemble models outperform state-of-the-art classifiers in terms of accuracy and rule count.
STYLE-ANALYZER fixes code style inconsistencies without manual intervention.
problem Manual code reviews are time-consuming and error-prone.
method Unsupervised machine learning with decision tree forest model.
result STYLE-ANALYZER accurately fixes code formatting violations with interpretable rules.
A novel memory mechanism for reinforcement learning agents that stores past events in human-readable language.
problem Lack of interpretability in reinforcement learning agent's memory mechanisms.
method Uses CLIP to associate visual inputs with language tokens, then feeds these tokens to a pretrained language model.
result Significantly faster convergence on challenging continuous recognition tasks.
Automatically extracts hyperparameter schemas from AI library documentation.
problem Lack of machine-readable hyperparameter schemas for AI automation tools.
method Automatically mines Python docstrings in AI libraries to extract JSON Schemas.
result Effective at extracting machine-readable schemas from 119 AI models.
Computer-generated proofs led to a mathematical result.
problem Discovering a mathematical result through computer-generated proofs.
method Combining computer-generated, human-readable proofs with mathematical abstraction.
result Abstracted lemma leading to an interesting mathematical result.
IdBench benchmarks semantic representations of identifiers, revealing strengths and weaknesses.
problem Evaluating semantic representations of identifiers in source code.
method Created a benchmark using developer ratings, evaluated natural language and source code embeddings, and compared lexical string distance functions.
result No single technique provides a satisfactory representation of semantic similarities, but ensemble models can improve performance.
Shai-am simplifies ML for finance, solving code structure and scalability issues.
problem Challenges in integrating ML for investment strategies, including code structure and scalability.
method Integrates a Python framework with modern open-source technologies to manage containerized pipelines and unified interfaces.
result Facilitates collaborative work in quantitative finance by enhancing reusability and readability.
MRCpy implements minimax risk classifiers with performance guarantees and distribution shift adaptability.
problem Classical risk minimization approaches are not robust to distribution shifts.
method Robust risk minimization approach for minimax risk classifiers.
result MRCs provide performance guarantees and adapt to distribution shifts.
Torch-Struct simplifies structured prediction for deep learning.
problem Difficulty in integrating structured prediction algorithms with deep learning frameworks.
method Develops a library (Torch-Struct) that integrates structured prediction with vectorized, auto-differentiation-based frameworks.
result Significant performance gains over fast baselines and cross-algorithm efficiency.
Team QCRI-MIT detects hyperpartisan news with 72.9% accuracy.
problem Detecting hyperpartisan news from biased political content.
method Logistic regression model using engineered features from propaganda detection.
result Significant performance improvements with better feature pre-processing.
We show that the action of the mapping class group on bordered Floer homology in the second to extremal spin^c-structure is faithful. This paper is designed partly as an introduction to the subject, and much of it should be readable without a background in Floer homology.
The six nondegeneracy conditions of geometric nature that are satisfied by the only six possibly existing nondegenerate general classes I, II, III-1, III-2, IV-1, IV-2 of 5-dimensional CR manifolds are shown to be readable instantaneously from their elementarily normalized respective defining graphed equations, without…
Solves the challenge of retrieving item-specific financial information from Form 10-Q filings.
problem Retrieving item-specific information from Form 10-Q filings with varying formats and machine-readable hierarchy.
method Complements a rule-based algorithm with a Convolutional Neural Network (CNN) image classifier to itemize 10-Q files.
result Demonstrates a generalized pipeline for rapid data retrieval from a large volume of textual data.
Developed Taylor series for muscle-finger system analysis.
problem Understanding the complex relationship between muscle activity and finger movement.
method Used Dendrite Net to develop Taylor series and construct relation spectrum.
result Found muscle synergy and coupling in hand movement.
Generative Map learns interpretable neural network maps for camera localization.
problem Creating interpretable maps for neural network-based camera localization.
method Combining generative models with Kalman filters and incorporating additional sensor information.
result Generative Map predicts images closely resembling the true scene and achieves comparable localization performance.
Given a Morse function f on a closed manifold M with distinct critical values, and given a field F, there is a canonical complex, called the Morse-Barannikov complex, which is equivalent to any Morse complex associated with f and whose form is simple. In particular, the homology of M with coefficients in F is immediate…
We present the mathematical background of a software package that computes triangulations of mapping tori of surface homeomorphisms, suitable for Jeff Weeks's program SnapPea. It consists of two programs. jmt computes triangulations and prints them in a human-readable format. jsnap converts this format into SnapPea's t…
Graphs help study singular points of algebraic surfaces.
problem Understanding singular points of algebraic surfaces.
method Introduced special kinds of graphs and developed a calculus with graphs.
result Graphs provide a new way to classify and study singular points.
This article demonstrates that convolutional operation can be converted to matrix multiplication, which has the same calculation way with fully connected layer. The article is helpful for the beginners of the neural network to understand how fully connected layer and the convolutional layer work in the backend. To be c…
This paper proposes a framework to learn explainable rules from knowledge graphs for better recommendation.
problem Combining side information with explainability in recommendation systems.
method Joint learning framework integrating rule induction from knowledge graphs with a rule-guided neural recommendation model.
result Significant improvements in item recommendation performance over baselines.
The paper explains deep neural network predictions using logical proxies.
problem Creating understandable explanations for deep neural network predictions.
method Randomized propositionalization and Bayes-like approach to identify logical proxies.
result Models in first-order logic can approximate DRM's predictions in local regions.
We consider the problem of finding the minimizer of a function f:Rd→R of the finite-sum form minf(w)=1/n∑infi(w). This problem has been studied intensively in recent years in the field of machine learning (ML). One promising approach for large-scale data is to use a stoc…
This paper proposes an organized generalization of Newman and Girvan's modularity measure for graph clustering. Optimized via a deterministic annealing scheme, this measure produces topologically ordered graph clusterings that lead to faithful and readable graph representations based on clustering induced graphs. Topog…
USS fund risk assessment shows low default chance but high overfunding.
problem Risk assessment of Universities Superannuation Scheme (USS) fund.
method Estimates risk of default and overfunding using a cautious model.
result Fund has less than 7% chance of defaulting but overfunding by at least £100bn.
BTZSC benchmarks zero-shot text classification across diverse models.
problem Systematically comparing zero-shot text classification across various models.
method Comprehensive benchmark of 22 datasets, comparing NLI cross-encoders, embedding models, rerankers, and instruction-tuned LLMs.
result Rerankers and instruction-tuned LLMs outperform NLI cross-encoders, with rerankers setting a new state-of-the-art.
Tracr compiles programs into transformer models for interpretability.
problem Uncertainty in understanding transformer model outputs due to unknown learned programs.
method Tracr compiles human-readable programs into known structure transformer models.
result Known structure of Tracr-compiled models serves as ground-truth for interpretability.
Paper provides a rigorous proof of the index theorem for economists.
problem Lack of a rigorous proof in textbooks for economists.
method Constructs a readable proof of the index theorem under specific assumptions.
result Provides a gap-free proof of the index theorem.
This is an introduction to the subject of the differential topology of the space of smooth loops in a finite dimensional manifold. It began as the background notes to a series of seminars given at NTNU and subsequently at Sheffield. I am posting them in the hope that they will be useful to people wishing to know a litt…
Project classifies malware using a multi-input LSTM model.
problem Classifying malware with high accuracy and handling class imbalance.
method Supervised multi-class LSTM model trained on annotated malware files.
result Multi-input model using both text and hexadecimal inputs outperformed single formats.
Paper tackles score following in full-page sheet music images.
problem Score following in unprocessed sheet music images.
method Directly predicts score positions from audio and image input.
result Outperforms state-of-the-art methods in alignment precision.
LLMs can help explain credit risk models but not autonomously.
problem Leveraging LLMs for post-hoc explainability in credit risk models.
method Comparison of LLM outputs with SHAP and coefficient-based attributions on three LMs.
result LLMs reliably preserve feature-importance rankings but poorly align with autonomous explanations.
A substantial progress in development of new and efficient tensor factorization techniques has led to an extensive research of their applicability in recommender systems field. Tensor-based recommender models push the boundaries of traditional collaborative filtering techniques by taking into account a multifaceted nat…
VALC provides concept-level interpretations of FLMs, overcoming word-level limitations.
problem Lack of higher-level structure interpretation in FLMs' attention weights.
method Formal definition of conceptual interpretation, variational Bayesian framework (VALC).
result VALC finds optimal language concepts for FLM predictions, providing concept-level interpretations.
System suggests clinical concepts in real-time for faster note creation.
problem Efficiently creating structured clinical notes with minimal keystrokes.
method Contextual autocompletion using shallow neural networks.
result Reduces keystrokes by 67% in real hospital environments.
OutlierTree detects outliers using decision trees and provides explanations.
problem Detecting outliers in data while providing understandable explanations.
method Supervised decision tree splits with 1-d confidence intervals.
result Human-readable explanations for outlier detection.
NFTs with diverse rare attributes sell at higher prices.
problem Understanding how rarity affects NFT market dynamics.
method Analyzed 3.7M NFT transactions across 410 collections.
result Rarer NFTs sell for higher prices and are less risky.
Researchers often summarize their work in the form of posters. Posters provide a coherent and efficient way to convey core ideas from scientific papers. Generating a good scientific poster, however, is a complex and time consuming cognitive task, since such posters need to be readable, informative, and visually aesthet…
Benchmark for math reasoning models from human proofs.
problem Measuring and accelerating machine learning models in high-level mathematical reasoning.
method Built a non-synthetic dataset from theorem prover proofs, defined a task for model to fill in missing propositions, used hierarchical transformer to improve performance.
result Neural models can capture non-trivial mathematical reasoning, hierarchical transformer outperforms baseline.
Develops deep learning for logical code segmentation.
problem Lack of logically segmented source code.
method Novel deep learning approach to generate logical code segments.
result Improves software analysis tasks like commenting, bug detection, and code synthesis.
NCVis speeds up data visualization for large datasets.
problem Performance issues in t-SNE for large datasets.
method Noise contrastive estimation for scalable visualization.
result NCVis outperforms state-of-the-art techniques in speed and quality.
Text clustering method replaces centroids with summaries for interpretability and scalability.
problem Efficiently clustering text data while maintaining interpretability and scalability.
method k-NLPmeans and k-LLMmeans, which periodically replace numeric centroids with textual summaries.
result Consistently outperforms classical baselines and recent LLM-based clustering methods.
Discover governing equations from data without specifying terms.
problem Discovering differential equations from data without predefined terms.
method Data-driven approach using genetic programming and automatic differentiation.
result Calibrated differential equations from various solutions of a differential equation.