Daily Papers of 2024-04-08

1. Direct Nash Optimization: Teaching Language Models to Self-Improve with General Preferences 49 upvotes, #1 of 2024-04-08
2. No "Zero-Shot" Without Exponential Data: Pretraining Concept Frequency Determines Multimodal Model Performance 24 upvotes, #2 of 2024-04-08
3. Stream of Search (SoS): Learning to Search in Language 18 upvotes, #3 of 2024-04-08
4. CantTalkAboutThis: Aligning Language Models to Stay on Topic in Dialogues 17 upvotes, #4 of 2024-04-08
5. RL for Consistency Models: Faster Reward Guided Text-to-Image Generation 13 upvotes, #5 of 2024-04-08
6. Social Skill Training with Large Language Models 12 upvotes, #6 of 2024-04-08
7. Chinese Tiny LLM: Pretraining a Chinese-Centric Large Language Model 8 upvotes, #7 of 2024-04-08
8. Robust Gaussian Splatting 7 upvotes, #8 of 2024-04-08
9. Sigma: Siamese Mamba Network for Multi-Modal Semantic Segmentation 5 upvotes, #9 of 2024-04-08

Data: hysts-bot-data/daily-papers-stats and the Daily Papers API. Open data: tardellirs/paper-pulse-data. Sister project: Model Pulse, the download history of every model on the Hub.