LLMs
Latest breakthroughs in Large Language Models
Articles
Large Language Model-Assisted Preparation of Transportation⭐9
This study explores LLM-assisted generation of Transportation Management Plans using the WisDOT WisTMP system, showing improved efficiency but remaining challen...
Wieszcz-XIX: A 3.1-Billion-Word Corpus of Pre-1918 Polish and⭐9
Wieszcz-XIX offers a 3.1-billion-word corpus of pre-1918 Polish and temporally bounded language models trained from scratch, with quality audits and OCR error a...
Agent-Controlled Forgetting: Reversible Context Curation for⭐10
Tool-using AI agents can reversibly archive verbose tool results, replacing them with short notes to cut prompt tokens and API costs by up to 50% in noisy traje...
Synthesis Through Simulation: Generating Coherent Enterprise⭐9
STS generates coherent enterprise data without schemas via LLM agents in simulated policy-enforcing environments, achieving 0.88 fidelity and 100% constraint sa...
Diffu-LoRA: A Novel Low-Rank Adaptation for Personalized⭐8
Diffu-LoRA introduces gated low-rank adaptation with bilevel optimization and progressive pruning to personalize diffusion models efficiently while preserving s...
An Explainable Header-Centric Framework for Large-Scale Semantic⭐8
An explainable header-centric framework for metadata-only semantic table interpretation maps headers to 39 types, detects data quality issues, and evaluates 120...
Adaptive Workflow Intelligence: A Cognitive Architecture for⭐9
Adaptive Workflow Intelligence (AWI) is a cognitive architecture using a Perception–Cognition–Action–Reflection loop to enable context-driven enterprise automat...
Tokka-Bench: A Standardized Framework for Evaluating Tokenizers⭐9
Tokka-Bench benchmarks seven BPE tokenizers across 100 natural and 20 programming languages, finding vocabulary allocation matters more than size.
GAMEGO: Training Game-Dev Agents with Synthetic Trajectories⭐8
GameGo transforms brief game seeds into detailed product specs, training coding agents on 55,060 synthetic trajectories to generate complex 2D, 2.5D, and 3D gam...
Zero-Shot Visualization: Exploring Text Corpora with User⭐9
Explore zero-shot visualization with LLMs to map text corpora onto user-prompted concept axes, comparing scoring methods for semantic faithfulness, fidelity, an...