HAGE: Advanced RL-Based Memory Graph for AI Models
Discover HAGE, an RL-driven weighted graph framework enhancing memory retrieval in AI for improved reasoning and dynamic relational understanding.
EXPO: Adaptive Policy Optimization for AI Exploration
Discover EXPO, a novel reinforcement learning method improving AI exploration via adaptive KL regulation and Gaussian curriculum sampling for better math r...
RADAR: Efficient Multi-Agent Communication Structure Generation
Discover RADAR, a redundancy-aware diffusion framework that optimizes multi-agent communication for higher accuracy and lower token use in AI systems.
Mitigating Cross-Modal Interference in Audio-Visual LLMs
Discover how the Separate First, Fuse Later framework improves audio-visual LLMs by reducing cross-modal interference and boosting reasoning accuracy.
KnotBench: Challenging Vision-Language Models with Knot Reasoning
Discover KnotBench, a tough benchmark testing vision-language models on complex knot diagram reasoning and operational understanding.