ResearchFastino Research·193d ago·3 min readSummaryListen 0:44
Correcting Stochastic Update Bias in Preconditioned Language Model OptimizersMethod corrects stochastic update bias in preconditioned language model optimizers.
News Hub
What actually shipped in agent engineering, pulled from the labs, arXiv and Hacker News.
See who we follow7 articles
ResearchFastino Research·193d ago·3 min readSummaryListen 0:43
GLiNER2-PII: A Multilingual Model for Personally Identifiable Information ExtractionGLiNER2-PII extracts personally identifiable information across multiple languages.
ResearchFastino Research·193d ago·3 min readSummaryListen 0:51
GLiGuard: Schema-Conditioned Classification for LLM Content ModerationGLiGuard uses schema-conditioned classification to moderate LLM-generated content.
ResearchFastino Research·223d ago·3 min readSummaryListen 1:00
Pioneer Agent: Continual Improvement of Small Language Models in ProductionPioneer Agent continuously improves small language models deployed in production.
ResearchFastino Research·405d ago·3 min readSummaryListen 0:48
Beyond Reactivity: Measuring Proactive Problem solving in LLM AgentsFramework measures proactive problem-solving capabilities in LLM agents beyond reactive behavior.
ResearchFastino Research·497d ago·3 min readSummaryListen 0:42
GLiNER2: An Efficient Multi-Task Information Extraction System with Schema-Driven InterfaceGLiNER2 performs multi-task information extraction using schema-driven interface design.
ResearchFastino Research·872d ago·3 min read
GLiNER: Generalist Model for Named Entity Recognition using Bidirectional TransformerGLiNER is a generalist model for named entity recognition using bidirectional transformers.
The Agent Architect
One pattern, one tradeoff, one production failure story. A short weekly briefing for people building agentic systems.
Weekly email, one-click unsubscribe. We only use your address to send the briefing.