<![CDATA[Stories by Samuel Flender on Medium]]> https://medium.com/@samuel.flender?source=rss-ce56d9dcd568------2 https://cdn-images-1.medium.com/fit/c/150/150/1*IHdofFy_lqh1Dpw4uhZkcw.jpeg Stories by Samuel Flender on Medium https://medium.com/@samuel.flender?source=rss-ce56d9dcd568------2 Medium Sat, 14 Nov 2026 04:53:20 GMT <![CDATA[Machine Learning in Fraud Detection: A Primer]]>

Balancing automation, accuracy, and customer experience in an ever-evolving adversarial landscape

Continue reading on TDS Archive »

]]> https://medium.com/data-science/machine-learning-in-fraud-detection-a-primer-8005b8c88cde?source=rss-ce56d9dcd568------2 https://medium.com/p/8005b8c88cde Tue, 12 Nov 2024 01:11:59 GMT 2024-11-14T05:59:31.480Z <![CDATA[User Action Sequence Modeling: From Attention to Transformers and Beyond]]>

The quest to LLM-ify recommender systems

Continue reading on TDS Archive »

]]> https://medium.com/data-science/user-action-sequence-modeling-from-attention-to-transformers-and-beyond-5f280268b399?source=rss-ce56d9dcd568------2 https://medium.com/p/5f280268b399 Mon, 15 Jul 2024 22:39:07 GMT 2024-07-15T22:39:07.736Z <![CDATA[LoRA: Revolutionizing Large Language Model Adaptation without Fine-Tuning]]>

Exploiting the low-rank nature of weight updates during fine-tuning results in orders of magnitude reduction in learnable parameters

Continue reading on TDS Archive »

]]> https://medium.com/data-science/lora-revolutionizing-large-language-model-adaptation-without-fine-tuning-3279d909f5af?source=rss-ce56d9dcd568------2 https://medium.com/p/3279d909f5af Tue, 23 Apr 2024 14:17:05 GMT 2024-04-23T14:17:05.062Z <![CDATA[Demystifying Mixtral of Experts]]>

Mistral AI’s open-source Mixtral 8x7B model made a lot of waves — here’s what’s under the hood

Continue reading on TDS Archive »

]]> https://medium.com/data-science/demystifying-mixtral-of-experts-498fe3b9bcf4?source=rss-ce56d9dcd568------2 https://medium.com/p/498fe3b9bcf4 Sun, 17 Mar 2024 14:05:07 GMT 2024-03-17T14:05:07.547Z <![CDATA[The Rise of Sparse Mixtures of Experts: Switch Transformers]]>

A deep-dive into the technology that paved the way for the most capable LLMs in the industry today

Continue reading on TDS Archive »

]]> https://medium.com/data-science/the-rise-of-sparse-mixtures-of-experts-switch-transformers-39cf3671f8a0?source=rss-ce56d9dcd568------2 https://medium.com/p/39cf3671f8a0 Thu, 15 Feb 2024 18:16:56 GMT 2024-02-15T18:16:56.625Z <![CDATA[Pushing the Limits of the Two-Tower Model]]>

Where the assumptions behind the two-tower model architecture break — and how to go beyond

Continue reading on TDS Archive »

]]> https://medium.com/data-science/pushing-the-limits-of-the-two-tower-model-a577090e5140?source=rss-ce56d9dcd568------2 https://medium.com/p/a577090e5140 Sun, 10 Dec 2023 15:26:19 GMT 2024-02-14T04:39:10.954Z <![CDATA[Towards Understanding the Mixtures of Experts Model]]>

New research reveals what happens under the hood when we train MoE models

Continue reading on TDS Archive »

]]> https://medium.com/data-science/towards-understanding-the-mixtures-of-experts-model-45d11ee5d50d?source=rss-ce56d9dcd568------2 https://medium.com/p/45d11ee5d50d Tue, 14 Nov 2023 15:23:54 GMT 2023-11-14T15:23:54.177Z <![CDATA[The Rise of Two-Tower Models in Recommender Systems]]>

A deep-dive into the latest technology used to debias ranking models

Continue reading on TDS Archive »

]]> https://medium.com/data-science/the-rise-of-two-tower-models-in-recommender-systems-be6217494831?source=rss-ce56d9dcd568------2 https://medium.com/p/be6217494831 Sun, 29 Oct 2023 15:33:11 GMT 2023-10-29T16:30:31.124Z <![CDATA[The Multi-Task Optimization Controversy]]>

Do we need special algorithms to train models on multiple tasks at the same time?

Continue reading on TDS Archive »

]]> https://medium.com/data-science/the-multi-task-optimization-controversy-793cbb431d98?source=rss-ce56d9dcd568------2 https://medium.com/p/793cbb431d98 Fri, 29 Sep 2023 04:21:12 GMT 2023-09-29T07:22:31.096Z <![CDATA[Machine Learning with Expert Models: A Primer]]>

How a decades-old idea enables training outrageously large neural networks today

Continue reading on TDS Archive »

]]> https://medium.com/data-science/machine-learning-with-expert-models-a-primer-6c74585f223f?source=rss-ce56d9dcd568------2 https://medium.com/p/6c74585f223f Tue, 05 Sep 2023 22:37:21 GMT 2023-09-09T03:22:51.571Z