All Articles
18316 articles total
Towards Engineering Scaling Laws with Pretraining Data Composition
Researchers show that pretraining data composition can be engineered to influence scaling laws in particle physics models.
Cross-Dataset, Age, and Gender Generalization: A Comprehensive Analysis of Fine-Tuning Strategies for Low-Resource Children's ASR
A study investigates fine-tuning strategies to improve Automatic Speech Recognition (ASR) for children with dysarthric speech.
Systematic Study of Dysarthric Speech Recognition: Spectral Features and Acoustic Models
A systematic study of spectral features and acoustic models for recognizing dysarthric speech in low-resource settings.
"No Feigning Surprise"
A Hacker News discussion thread with no specific technical content provided in the snippet.
Rocket Report: Rebuild begins at Blue Origin launch pad; Relativity targets Mars
News regarding Blue Origin's launch pad rebuild and Relativity Space's targets for Mars, including a French startup changing its rocket name.
FineREX: Fine-Tuned NER-RE for Human Smuggling Knowledge Graphs
Introduction of FineREX, a pipeline using fine-tuned LLMs for named entity recognition and relationship extraction to build knowledge graphs from legal documents.
AURA: Adaptive Uncertainty-aware Refinement for LLM-as-a-Judge Auditing
AURA is a new adaptive uncertainty-aware refinement framework designed to audit LLM-as-a-judge decisions using selective human verification.
OnDeFog: Online Decision Transformer under Frame Dropping
OnDeFog proposes integrating DeFog mechanisms with an online decision transformer to handle frame dropping in reinforcement learning environments.
Library-Aware Doubles and Iterative Repair for Large Language Model-Generated Unit Tests in OpenSIL Firmware
A multi-agent LLM pipeline for automated unit test generation and repair in the OpenSIL firmware codebase, achieving high line coverage for C firmware.
NRITYAM: Language Models Meet Art and Heritage of Dance
NRITYAM is a large-scale multilingual benchmark for evaluating LLM cultural comprehension of global dance traditions.
Bidirectional Tutoring for Developmental Motor Learning in Robots: Co-Developed Interaction Dynamics Support Stable Learning
Research on bidirectional tutoring for humanoid robots, using a free-energy-principle-based neural network to support stable motor skill learning.
VOiLA: Vectorized Online Planning with Learned Diffusion Model for POMDP Agents
VOiLA is a framework that learns task-agnostic POMDP models using diffusion models for efficient online planning in autonomous robots.
QueryGaussian: Scalable and Training-Free Open-Vocabulary 3D Instance Retrieval
QueryGaussian is a training-free framework for scalable open-vocabulary 3D instance retrieval in large-scale scenes using consumer-grade hardware.
The room the economy can't see
An exploration of the invisible parts of the economy that are not captured by traditional economic metrics.
Leave a Trace
A discussion or piece about the concept of leaving a trace or digital footprint.
Show HN: Modeloop – From visual algorithms to microcontroller C code
Modeloop is a tool that converts visual algorithms into C code for microcontrollers, simplifying embedded development.
The CEO of Allbirds’ new AI biz has a plan, but no employees
A sole founder starts an AI business with a significant seed round but currently has no employees.
T1 Phone PR firm is ‘not assisting Trump Mobile any further’
A PR firm has ceased assisting Trump Mobile following communication issues.
In season 2 of Sugar, Colin Farrell’s quirky detective becomes much more human
Details on the second season of the Apple TV show Sugar, featuring Colin Farrell.
Kaleidescape’s movie player blows streaming, and your wallet, away
A look at Kaleidescape's high-end movie player, which offers superior quality over streaming services.
