The world's leading publication for data science and artificial intelligence professionals. Submit an Article ✍️ https://xitter.cf/t.co/57pIMegK1o
We're Global 🌏
Joined October 2016
- Tweets64.6K
- Following1.9K
- Followers251K
- Likes8K
Towards Data Science retweeted
One benchmark: Claude Code used 33K tokens, while Cursor used 188K.
Another: the exact reverse.
Token cost is a property of the task, not the tool. The field test that actually settles it: towardsdatascience.com/claud…
@TDataScience
Irfan Alghani Khalid's Food Police reads a photo of his meal with the model built into macOS, Apple's Foundation Model, and logs every alert to a Google Sheet.
towardsdatascience.com/ai-fo…
For his latest exploration of agentic AI workflows, @BenjaminNweke11 zooms in on what happens when a tool breaks — but the dashboard stays green. towardsdatascience.com/agent…
"The models reason fine, but [...] the underlying enterprise data isn't accessible, clean, or structured enough for them to act on reliably."
Partha Sarkar walks us through six advanced architectural patterns for data practitioners. towardsdatascience.com/data-…
Not sure how to add guardrails to your LLM agents, or even whether you should? Shuai Guo unpacks the benefits, and outlines potential approaches to making your agentic systems more resilient. towardsdatascience.com/how-t…
Learn all about the new profiler included in Python 3.15 — @taupirho presents a detailed, hands-on guide to the profiling.sampling module, known as Tachyon. towardsdatascience.com/pytho…
Towards Data Science retweeted
Check out my latest article about unexpected costs when routing to a cheaper LLM on @TDataScience
#AgenticAI #Routing #LLM #ArtificialIntelligence
LLM routing is supposed to drive costs down without sacrificing efficiency.
@nhu_hoang's incisive deep dive shows how switching models mid-session can erase the savings promised by cheaper inference. towardsdatascience.com/the-h…
Towards Data Science retweeted
I’m three months into my transition into data engineering. And so far I’ve built projects that tested me on various skills in data engineering. Here’s my most recent installment.
I built a pipeline that tracks the job market for roles I’m trying to get hired for. Read it here.
"So I built a pipeline whose entire job is to watch the job market for the exact roles I'm trying to get hired into: data engineer, analytics engineer, ETL developer, and other related jobs."
@ibbysalam unpacks the process of building an incremental pipeline, and the trouble spots he encountered along the way. towardsdatascience.com/i-bui…
Towards Data Science retweeted
My ML pipeline cheated.@TDataScience
Cross-validation didn't notice.
The scores looked great — because my model had access to information it should never have seen.
I fixed the pipeline and reran everything.
The result was painful.
towardsdatascience.com/my-ma…
Towards Data Science retweeted
My latest article on Python 3.15's new profiler is now available to read on @TDataScience.
track.pstmrk.it/3s/towardsda…
"Instead of looking for suspicious vocabulary, repetitive sentence structures, or everyone’s favorite AI punctuation mark (and mine!), the em dash, the authors investigated the narrative decisions that shape a story."
Sara A. Metwalli explores recent research of the kinds of stories AI tools generate. towardsdatascience.com/what-…
Even if neural networks won't settle the measurement problem at the center of Schrödinger’s cat, Samit Ganguly suggests they're already changing the way we see, predict, and protect quantum superpositions. towardsdatascience.com/openi…
How do you check an app a coding agent wrote if you can't read the code?
Reya Vir's FlowCheck injects IDs into every element.
towardsdatascience.com/flowc…
How many independent answers can you actually get when you ask several models the same question?
@mangal_utkarsh explains why the answer is often a smaller number than you'd expect. towardsdatascience.com/your-…
What does effective dimension reveal about model redundancy, generalization, and trainability that parameter count, loss, and accuracy do not?
Adrian Panasiewicz and Jacob Cybulski launch a new series unpacking the factors that shape model outputs and performance. towardsdatascience.com/why-p…
Taking your first steps in data science? If you could use an accessible, detailed explainer on skewness and kurtosis, we recently shared Mansi Bramta's comprehensive guide. towardsdatascience.com/skewn…
"Robust estimators address this problem by limiting the influence of observations that are inconsistent with the fitted relationship."
Messy data? Don't let it harm the accuracy of your curves — Aamir Hussain Chughtai explains how to address this common challenge. towardsdatascience.com/dont-…
On-policy self-distillation has taken ML research by storm in 2026. But does it hold up?
Jonathan Yahav looks at the promising results on continual learning, and at the open problems: loss of output diversity and trouble with thinking models. towardsdatascience.com/could…
What happens when we try to lead small language models astray by presenting them with wrong hints in front of exam questions?
@mangal_utkarsh shares insights and practical takeaways based on his extensive experiments. towardsdatascience.com/same-…
Can you watch one crystal phase turn into another as it happens?
@NCocaLopez shows how in situ Raman spectroscopy and a simple unmixing model track the FAU → GIS zeolite transformation during synthesis. towardsdatascience.com/raman…
