ResearchResearch paperLarge Language Models · Reinforcement Learning · Data & Datasets2 sources · Oct 10, 2026

Vela: User Behavior as Supervision Sparks LLM Intelligence

We propose user behavior as supervision and introduce Vela, a framework that converts real user behavior data into trainable tasks and trajectories.

Large language models are increasingly deployed in applications that act on users' long-term behavior histories, from personal assistants to recommendation and advertising systems.
These applications demand the user dimension of LLM intelligence: inferring a person's preferences, intent, and likely future actions from weeks to years of behavior data.
From the anonymized behavior records of 100,000 users, Vela generates 100,000 tasks across 91 real-world task templates; difficulty-based filtering and trajectory acceptance criteria retain approximately 75,000 tasks with selected training trajectories.
Training on more user behavior data further improves user-related performance, and reinforcement learning after fine-tuning outperforms reinforcement learning from the base model by up to 7.7 points.

[1]Vela: User Behavior as Supervision Sparks LLM Intelligence
Hugging Face Daily Papers · Oct 10, 12:00 AM
We propose user behavior as supervision and introduce Vela, a framework that converts real user behavior data into trainable tasks and trajectories.
Large language models are increasingly deployed in applications that act on users' long-term behavior histories, from personal assistants to recommendation and advertising systems.
[2]Vela: User Behavior as Supervision Sparks LLM Intelligence
arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 10, 08:28 AM · same content

Extractive summary: sentences quoted from the sources.

Before this