Daily Papers of 2025-09-08

1. Why Language Models Hallucinate 163 upvotes, #1 of 2025-09-08
2. Set Block Decoding is a Language Model Inference Accelerator 49 upvotes, #2 of 2025-09-08
3. Symbolic Graphics Programming with Large Language Models 44 upvotes, #3 of 2025-09-08
4. LuxDiT: Lighting Estimation with Video Diffusion Transformer 14 upvotes, #4 of 2025-09-08
5. WildScore: Benchmarking MLLMs in-the-Wild Symbolic Music Reasoning 11 upvotes, #5 of 2025-09-08
6. LatticeWorld: A Multimodal Large Language Model-Empowered Framework for Interactive Complex World Generation 9 upvotes, #6 of 2025-09-08
7. WinT3R: Window-Based Streaming Reconstruction with Camera Token Pool 6 upvotes, #7 of 2025-09-08
8. Behavioral Fingerprinting of Large Language Models 5 upvotes, #8 of 2025-09-08
9. Bootstrapping Task Spaces for Self-Improvement 5 upvotes, #8 of 2025-09-08
10. On Robustness and Reliability of Benchmark-Based Evaluation of LLMs 4 upvotes, #10 of 2025-09-08
11. U-ARM : Ultra low-cost general teleoperation interface for robot manipulation 3 upvotes, #11 of 2025-09-08
12. MedVista3D: Vision-Language Modeling for Reducing Diagnostic Errors in 3D CT Disease Detection, Understanding and Reporting 3 upvotes, #11 of 2025-09-08

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.