1Merge and Conquer: Evolutionarily Optimizing AI for 2048 (opens in new tab)(arxiv.org)arXiv1xianshou8mo ago0Save
2Stuck in the Matrix: Probing Spatial Reasoning in Large Language Models (opens in new tab)(arxiv.org)arXiv1xianshou8mo ago0Save
3Reflection AI Raises $2B to Build "American DeepSeek" (opens in new tab)(nytimes.com)9xianshou8mo ago2Save
4Nvidia-backed Reflection AI raising at $5.5B valuation (opens in new tab)(reuters.com)2xianshou8mo ago1Save
6DeepSeek V3 0324 is now the best nonthinking model (Reddit) (opens in new tab)(old.reddit.com)1xianshou1y ago0Save
7DeepSeek V3 0324 outpaces GPT 4.5 and Claude 3.7 in coding, other benchmarks (opens in new tab)(huggingface.co)7xianshou1y ago0Save
9InvestorBench: A Benchmark for Financial Decision-Making Tasks with Agents (opens in new tab)(arxiv.org)arXiv1xianshou1y ago0Save
10An Evolved Universal Transformer Memory (Sakana.ai) (opens in new tab)(sakana.ai)1xianshou1y ago0Save
