How Lightweight Large Language Models Perform in Biomedical Named Entity Recognition
A new study systematically evaluates the performance of lightweight large language models on biomedical named entity rec…
📂 Research
Latest articles in Research
SciHorizon-DataEVA: An AI Agent That Automatically Evaluates the AI-Readiness of Scientific Data
A latest arXiv paper introduces the SciHorizon-DataEVA system, which leverages an agentic architecture to automatically …
FutureWorld: Training Predictive Agents with Real-World Outcomes
A research team has introduced FutureWorld, a live environment platform designed for real-time future prediction tasks. …
Safety Stress Test for 72 LLMs: Who Dares Let Them Control Medical Care Robots?
A new study constructs a dataset of 270 harmful instructions based on AMA ethical guidelines to benchmark 72 large langu…
New LLM Benchmarking Framework: Human-AI Collaborative Assessment of Mathematical Competency
A new study proposes a 'human-in-the-loop' benchmarking framework that systematically evaluates the performance of multi…
AGEL-Comp: A Neuro-Symbolic Framework Tackling the Compositional Generalization Challenge for AI Agents
A latest arXiv paper introduces AGEL-Comp, a neuro-symbolic AI agent architecture featuring three core innovations inclu…
Symbol Grounding ≠ Compositional Reasoning: Core Assumption of Neuro-Symbolic Systems Challenged by Empirical Evidence
A latest arXiv study presents the first systematic empirical analysis showing that the core assumption in neuro-symbolic…
Breaking Through LLM Reasoning Bottlenecks: Auto-Relational Reasoning Framework Emerges
A latest arXiv paper proposes a theoretical framework called "Auto-Relational Reasoning," aiming to break through the so…
Real Money, Real Stakes: A Study on Operating-Layer Controls for Onchain LLM Agents
A latest arXiv paper investigates how autonomous language model agents can reliably translate user instructions into onc…
DreamProver: Revolutionizing Theorem Proving with the Wake-Sleep Paradigm
Researchers propose the DreamProver framework, which leverages the "Wake-Sleep" program induction paradigm to automatica…
Study Reveals 'Persuadability' Risks of LLMs as Legal Adjudication Tools
A new study examines the 'persuadability' problem of large language models in legal decision-making scenarios, finding t…
OMEGA Framework: Using AI to Automatically Generate and Optimize Machine Learning Algorithms
A research team has introduced OMEGA, an end-to-end automated AI research framework that automatically creates novel ML …