MOLE: Metadata Extraction and Validation in Scientific Papers Using LLMs Paper • 2505.19800 • Published 2 days ago • 1
An Embarrassingly Simple Defense Against LLM Abliteration Attacks Paper • 2505.19056 • Published 3 days ago • 3
Beyond the Last Answer: Your Reasoning Trace Uncovers More than You Think Paper • 2504.20708 • Published 29 days ago • 22
Towards Data-Efficient Pretraining for Atomic Property Prediction Paper • 2502.11085 • Published Feb 16 • 3
Over-Tokenized Transformer: Vocabulary is Generally Worth Scaling Paper • 2501.16975 • Published Jan 28 • 31
Model Merging and Safety Alignment: One Bad Model Spoils the Bunch Paper • 2406.14563 • Published Jun 20, 2024 • 31
SynthCLIP: Are We Ready for a Fully Synthetic CLIP Training? Paper • 2402.01832 • Published Feb 2, 2024 • 6
Mindstorms in Natural Language-Based Societies of Mind Paper • 2305.17066 • Published May 26, 2023 • 3