Tag: AI language models
Browse our exclusive articles!
ORBIT: Budget-Friendly Scalable Data for Search Agents
Discover ORBIT, a scalable, verifiable dataset for training search agents with multi-step reasoning, all created without costly API services.
Multiscreen: Efficient Attention with Absolute Relevance
Discover Multiscreen, a novel attention mechanism enhancing NLP models by screening irrelevant keys for better performance and efficiency.
Backdoor Attacks on Continuous Latent Reasoning Models
Explore how backdoor attacks exploit continuous latent reasoning in AI models, revealing vulnerabilities and evading defenses with the ThoughtSteer method.
Scaling Laws for Optimizing RAG-Based Language Models
Explore how scaling pretraining and retrieval improves RAG language models for knowledge-intensive tasks with optimal data allocation strategies.
Diversity-Aware Reverse KL Divergence for LLM Distillation
Explore Diversity-aware Reverse KL Divergence (DRKL) to improve large language model distillation with better performance and output diversity.