Papers

Filtered to capability heads · clear filter

Browse by term

continual learning 64reinforcement learning 34large language models 17benchmarking 12vision-language models 10generative models 8language models 8video generation 7multimodal models 6natural language processing 6robotics 6world models 6benchmarks 5diffusion models 5on-policy distillation 5policy optimization 5scalability 5self-distillation 5vision-language-action models 5autoregressive models 4computer vision 4diffusion transformers 4LLMs 4multimodal large language models 4verifiable rewards 4attention mechanisms 3embodied intelligence 3image editing 3long-term memory 3multimodal learning 3

Matching papers

Multi-Head Latent Control: A Unified Interface for LLM Agent Decision Making

7 upvotes · 15 JUL 2026 · Amirhosein Ghasemabadi, Ruichen Chen, Bahador Rashidi et al.

This paper introduces a new method, Multi-Head Latent Control, that allows large language models to make decisions at inference time, such as whether to use a stronger model or request additional information, without relying on costly and difficult-to-maintain external signals. Practitioners can use this method to improve the quality and efficiency of multi-model systems.

latent controlLLMVLMdeployment-time control signalscapability headsresolution heads