KI-Papiere
Grundlegende Konzepte & Architekturen:
- Rückpropagation
- Faltende neuronale Netze
- Word2Vec-Papiere von Tomáš Mikolov.
- Sequenz-zu-Sequenz-Lernen mit neuronalen Netzen
- Attention is All You Need
- ResNet, “Deep Residual Learning for Image Recognition”
Große Sprachmodelle & verwandte Techniken:
- Chain-of-Thought Prompting Elicits Reasoning in Large Language Models
- Retrieval-Augmented Generation for Knowledge-Intensive NLP Tasks
- GPT-4 Technical Report
- Claude 3 Model Card
- LLaMA 3 Paper
Spezifische Modelle & Anwendungen:
- DeepSeek V3 & V2
- Whisper, Robust Speech Recognition via Large-Scale Weak Supervision
- Latent Diffusion Models (Stable Diffusion) Paper
- DALL-E 3 Scaling Text-to-Image Generation
Benchmarks & Evaluations:
- SWE-Bench Can Language Models Resolve Real-World GitHub Issues?
Kuratierte Listen:
- NeurIPS Test of Time Papers
- Ilyas Top 30 AI-Papiere Eine kuratierte Liste von Ilya Sutskever, verfügbar unter https://aman.ai/primers/ai/top-30-papers/.