The field of artificial intelligence is witnessing significant advancements in the development of large language models (LLMs) and agentic systems. Recent research has focused on enhancing the capabilities of LLMs, enabling them to interact with their environment, reason, and make decisions autonomously. The integration of tools and external knowledge has emerged as a key area of research, with studies demonstrating the benefits of in-tool learning over in-weight learning for factual recall. Furthermore, the development of agentic frameworks and architectures has enabled the creation of more sophisticated and adaptive systems. Noteworthy papers in this area include 'IR-Agent' and 'rStar2-Agent', which demonstrate the potential of LLMs in molecular structure elucidation and math reasoning, respectively. Overall, the field is moving towards the development of more advanced and autonomous systems, with significant implications for various applications and industries.
Advancements in Large Language Models and Agentic Systems
Sources
IR-Agent: Expert-Inspired LLM Agents for Structure Elucidation from Infrared Spectra
AgentScope 1.0: A Developer-Centric Framework for Building Agentic Applications
From Linear to Hierarchical: Evolving Tree-structured Thoughts for Efficient Alpha Mining
Adaptive Command: Real-Time Policy Adjustment via Language Models in StarCraft II
TextOnly: A Unified Function Portal for Text-Related Functions on Smartphones
Retrieval Capabilities of Large Language Models Scale with Pretraining FLOPs
From Language to Action: A Review of Large Language Models as Autonomous Agents and Tool Users
LLM-based Agentic Reasoning Frameworks: A Survey from Methods to Scenarios
TradingGroup: A Multi-Agent Trading System with Self-Reflection and Data-Synthesis
DiscussLLM: Teaching Large Language Models When to Speak
Technology-assisted Personalized Yoga for Better Health - Challenges and Outlook
AppAgent-Pro: A Proactive GUI Agent System for Multidomain Information Integration and User Assistance
Understanding Tool-Integrated Reasoning
Floor sensors are cheap and easy to use! A Nihon Buyo Case Study
Learning Game-Playing Agents with Generative Code Optimization
The Anatomy of a Personal Health Agent
AWorld: Orchestrating the Training Recipe for Agentic AI
rStar2-Agent: Agentic Reasoning Technical Report
Provable Benefits of In-Tool Learning for Large Language Models
ProactiveEval: A Unified Evaluation Framework for Proactive Dialogue Agents