LLMs
Latest breakthroughs in Large Language Models
Articles
CaVe-VLM-CoT: An Interpretable Vision-Language Model Framework⭐7
CaVe-VLM-CoT introduces an interpretable vision-language model framework using evidence-grounded reasoning to reduce hallucinations and improve trust in AI outp...
NAVI-Orbital: First In-Orbit Demonstration of a Zero-Shot⭐7
NAVI-Orbital demonstrates the first in-orbit zero-shot vision-language model for autonomous Earth observation, enabling natural language queries and
When Rules Learn: A Self-Evolving Agent for Legal Case Retrieval⭐8
A self-evolving LLM agent generates and refines query rewriting rules to enhance BM25 for legal case retrieval, outperforming static methods on LeCaRD-v2.
Beyond Parallel Sampling: Diverse Query Initialization for⭐10
Diverse query initialization (DivInit) boosts agentic search by replacing redundant parallel sampling, achieving 5-7% gains on multi-hop QA tasks.
Dr-DCI: Scaling Direct Corpus Interaction via Dynamic Workspace⭐9
Dr-DCI scales direct corpus interaction by dynamically expanding workspaces via retriever-steered actions, achieving 71.2% accuracy in Browsecomp-Plus benchmark...
A Definition of Good Explanations and the Challenges of⭐9
This article defines a "good explanation" using counterfactuals and prior beliefs, then discusses key challenges in applying this definition to large
A Deep Reinforcement Learning (DRL)-Based Transformer Method for⭐7
A Transformer model trained on small OSSP instances generalizes to 100x100 problems, rivaling heuristics with 12-15% optimality gaps.
Arbor: Tree Search as a Cognition Layer for Autonomous Agents⭐8
Arbor is a multi-agent framework using structured tree search as a cognition layer for autonomous agents, enabling full-stack LLM inference optimization
ToolSense: A Diagnostic Framework for Auditing Parametric Tool⭐8
ToolSense diagnoses whether LLMs truly understand tools in agent systems, revealing a 50-64% performance gap between benchmark and real-world retrieval.
Position: Hippocampal Explicit Memory Is the Cornerstone for AGI⭐8
A position paper argues hippocampal explicit memory is essential for AGI, as LLMs rely on implicit memory and lack higher-order reasoning like planning.