Daily Papers of 2024-02-08

1. Grandmaster-Level Chess Without Search 70 upvotes, #1 of 2024-02-08
2. BiLLM: Pushing the Limit of Post-Training Quantization for LLMs 51 upvotes, #2 of 2024-02-08
3. ScreenAI: A Vision-Language Model for UI and Infographics Understanding 45 upvotes, #3 of 2024-02-08
4. Direct Language Model Alignment from Online AI Feedback 36 upvotes, #4 of 2024-02-08
5. LGM: Large Multi-View Gaussian Model for High-Resolution 3D Content Creation 29 upvotes, #5 of 2024-02-08
6. ConsistI2V: Enhancing Visual Consistency for Image-to-Video Generation 27 upvotes, #6 of 2024-02-08
7. EfficientViT-SAM: Accelerated Segment Anything Model Without Performance Loss 24 upvotes, #7 of 2024-02-08
8. Hydragen: High-Throughput LLM Inference with Shared Prefixes 21 upvotes, #8 of 2024-02-08
9. The Hedgehog & the Porcupine: Expressive Linear Attentions with Softmax Mimicry 15 upvotes, #9 of 2024-02-08
10. CodeIt: Self-Improving Language Models with Prioritized Hindsight Replay 15 upvotes, #9 of 2024-02-08
11. Fine-Tuned Language Models Generate Stable Inorganic Materials as Text 9 upvotes, #11 of 2024-02-08
12. Fast Timing-Conditioned Latent Audio Diffusion 7 upvotes, #12 of 2024-02-08
13. TP-Aware Dequantization 4 upvotes, #13 of 2024-02-08
14. Progressive Gradient Flow for Robust N:M Sparsity Training in Transformers 2 upvotes, #14 of 2024-02-08

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.