Yannic Kilcher — transcripts.

1I BUILT A FULLY AUTOMATIC MANSPLAINER 2TiDAR: Think in Diffusion, Talk in Autoregression (Paper Analysis) 3Titans: Learning to Memorize at Test Time (Paper Analysis) 4[Paper Analysis] The Free Transformer (and some Variational Autoencoder stuff) 5[Video Response] What Cloudflare's code mode misses about MCP and tool calling 6[Paper Analysis] On the Theoretical Limitations of Embedding-Based Retrieval (Warning: Rant) 7AGI is not coming! 8Context Rot: How Increasing Input Tokens Impacts LLM Performance (Paper Analysis) 9Energy-Based Transformers are Scalable Learners and Thinkers (Paper Review) 10On the Biology of a Large Language Model (Part 2) 11On the Biology of a Large Language Model (Part 1) 12[GRPO Explained] DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models 13Byte Latent Transformer: Patches Scale Better Than Tokens (Paper Explained) 14Safety Alignment Should be Made More Than Just a Few Tokens Deep (Paper Explained) 15TokenFormer: Rethinking Transformer Scaling with Tokenized Model Parameters (Paper Explained) 16GSM-Symbolic: Understanding the Limitations of Mathematical Reasoning in Large Language Models 17Were RNNs All We Needed? (Paper Explained) 18Scaling LLM Test-Time Compute Optimally can be More Effective than Scaling Model Parameters (Paper) 19Privacy Backdoors: Stealing Data with Corrupted Pretrained Models (Paper Explained) 20Scalable MatMul-free Language Modeling (Paper Explained) 21Hallucination-Free? Assessing the Reliability of Leading AI Legal Research Tools (Paper Explained) 22xLSTM: Extended Long Short-Term Memory 23[ML News] OpenAI is in hot waters (GPT-4o, Ilya Leaving, Scarlett Johansson legal action) 24ORPO: Monolithic Preference Optimization without Reference Model (Paper Explained) 25[ML News] Chips, Robots, and Models 26TransformerFAM: Feedback attention is working memory 27[ML News] Devin exposed | NeurIPS track for high school students 28Leave No Context Behind: Efficient Infinite Context Transformers with Infini-attention 29[ML News] Llama 3 changes the game 30Hugging Face got hacked 31[ML News] Microsoft to spend 100 BILLION DOLLARS on supercomputer (& more industry news) 32[ML News] Jamba, CMD-R+, and other new models (yes, I know this is like a week behind 🙃) 33Flow Matching for Generative Modeling (Paper Explained) 34Beyond A*: Better Planning with Transformers via Search Dynamics Bootstrapping (Searchformer) 35[ML News] Grok-1 open-sourced | Nvidia GTC | OpenAI leaks model names | AI Act 36[ML News] Devin AI Software Engineer | GPT-4.5-Turbo LEAKED | US Gov't Report: Total Extinction 37No, Anthropic's Claude 3 is NOT sentient 38Gemini has a Diversity Problem 39V-JEPA: Revisiting Feature Prediction for Learning Visual Representations from Video (Explained) 40Lumiere: A Space-Time Diffusion Model for Video Generation (Paper Explained) 41AlphaGeometry: Solving olympiad geometry without human demonstrations (Paper Explained) 42Mixtral of Experts (Paper Explained) 43Until the Litter End 44LLaMA Pro: Progressive LLaMA with Block Expansion (Paper Explained) 45I created an AI-powered Social Network 46NeurIPS 2023 Poster Session 4 (Thursday Morning) 47Art @ NeurIPS 2023