about
Stories from February 4, 2026 (UTC)
Go back a day, month, or year. Go forward a day.
1. Attention at Constant Cost per Token via Symmetry-Aware Taylor Approximation (arxiv.org)
164 points by fheinsen 241 days ago | hide | past | pdf | 96 comments
2. Distance Marching for Generative Modeling (arxiv.org)
4 points by E-Reverance 241 days ago | hide | past | pdf | discuss
3. Who's in Charge? Disempowerment Patterns in Real-World LLM Usage (arxiv.org)
3 points by 7777777phil 241 days ago | hide | past | pdf | discuss
4. Recursive Knowledge Synthesis for Multi-LLM Systems (arxiv.org)
2 points by bob1029 241 days ago | hide | past | pdf | discuss
5. Pulse: Decentralized RL training centralized speed (100x weight sync reduction) (arxiv.org)
2 points by synapz_org 241 days ago | hide | past | pdf | discuss
6. Are Pre-Trained Convolutions Better Than Pre-Trained Transformers? (2021) (arxiv.org)
2 points by fzliu 242 days ago | hide | past | pdf | discuss
7. The Trigger in the Haystack: Extracting and Reconstructing LLM Backdoor Triggers (arxiv.org)
1 point by maltalex 241 days ago | hide | past | pdf | discuss
8. Pendulum: A Benchmark for Assessing Sycophancy in MLLM's (arxiv.org)
1 point by onestay42 242 days ago | hide | past | pdf | discuss