A personalized classifier needs several hundred labeled examples before its predictions become reliable.

Added to korrents 23 Jan 2026 · How quotes work · Something wrong? Tell us

Related korrents

Our reading — they may agree, disagree or merely touch the same thing. Closest first: a shared subject counts for most, then how near the wording is.

• Learned approximations of slow simulators change what science is possible, by turning a six-month screening run into something you do over lunch. JDJeff Dean30 Jul 2026similar wordingChief Scientist at Google DeepMind and Google Research
• Full fine-tuning of large language models has become too costly, making parameter-efficient fine-tuning methods necessary. SRSebastian Ruder5 Dec 2023similar wordingResearch scientist at Meta working on multilingual models and evaluation
• Using a neural network is likely to become easier than designing a custom-tuned prediction algorithm for many tasks. NENelson Elhage4 Jul 2025similar wordingSoftware engineer who writes the blog Made of Bugs about performance
• When pre-training data is limited, it is better to train a smaller model for multiple epochs than to train a larger model on unique data once. SRSebastian Ruder1 Dec 2023similar wording
• What sits in a model's context is far clearer to it than its training data, which is trillions of tokens stirred into a soup of parameters. JDJeff Dean30 Jul 2026similar wording

On the map

Open the map on its own page →