Sitemap - 2024 - Machine learning at scale

69. Closing the year off: stream-of-consciousness, looking back and forward

68. ColBERT and ColPALI: late interaction retrieval methods

67. Improving RAG components.

66. I am no longer postmortem free.

65. Finetuning LLMs to make them good at RAG: RankRAG by no other than NVIDIA.

64. Breaking the Attention Barrier: A Deep Dive into Scaling LLM Context Length

63. How multimodal LLMs (MLLM) work under the hood?

62. 7 Hard-Earned Lessons About Software Engineering Careers

61. How shortwave designed the world smartest email using RAG systems

60. How LinkedIn built its GenAI platform

59. LLM as a judge: use cases and evaluations.

58. My personal blueprint to get up and running fast when joining a new company / new team.

57. What happened to BERT?

56. Transfusion: Predict the Next Token and Diffuse Images with One Multi-Modal Model.

55. Artificial Intelligence Math Olympiad: Deep dive into Numina solution.

54. OpenAI o1 model: perspective from a ML system point of view.

53. Detecting hallucinations in large language models using semantic entropy.

52. Q*: Improving Multi-step reasoning with deliberative planning.

51. Short-circuiting LLMs to construct highly reliable safeguards

50. The Evolution of In-Context Learning: Implications for Machine Learning Engineers.

49. Autoregressive text-to-image vs Diffusion models.

48. High quality pre-training data: FineWeb 🍷

47. Feature stores in an embedding world

[Special edition] Scaling Your Impact: Lessons learned from 1,000 Pull Requests and 100,000 Lines of Code

46. Federated learning

45. Compound AI systems

44. Testing Machine Learning

43. Pick your own embedding dimension: Matryoshka representation learning (MRL)

42. Active learning

41. Semi-supervised learning

40. Automatically detecting Under-trained Tokens

39. Interpretable Features from Claude 3 Sonnet.

38. Training safety-first LLM models: Instruction Hierarchy from OpenAI.

37. Machine UN-Learning (!!)