All news

OpenAI explains how it built ads for ChatGPT around conversation context instead of discrete search events, with privacy and relevance controls.

Meta builds a stable token vocabulary for ads so its generative recommenders can reason over ad content without ID churn.

Pinterest releases ModuleLens, a tool to inspect PyTorch module structure and cost in large recommendation models, for engineers and agents.

Item-level adaptive recursion spends more compute on hard items in sequential recommendation.

Diffusion-guided contrastive learning for cross-domain sequential recommendation.

Information-bottleneck VAE with gating for top-N recommendation.

FINDUS detects disadvantaged user subgroups in recommender systems.

Diffusion generation for cold-start items in federated recommendation.

Combined popularity and positivity bias makes recommenders unfair.

LLM-enhanced GNNs for CTR prediction.

Survey and audit of recommender research where LLM agents edit recommender configs or code and retrain them.

Multimodal semantic fusion for video recommendation.

Mix-attention fuses multimodal LLM features into sequential recommendation.

Distills reliable recommendation rationales from LLMs for conversational recommenders.

Cost-efficient conversational recommendation using retrospective learning and isolated reflection.