Alonso Silva
@alonsosilva.bsky.social
Researcher on Verifiable AI @ Nokia Bell Labs. Interested on Large Language Models (LLMs), Machine Learning and Data Analysis.
Featured in marimo newsletter 🤩
marimo.io/blog/newslet...
marimo.io/blog/newslet...
September 29, 2025 at 6:14 PM
Featured in marimo newsletter 🤩
marimo.io/blog/newslet...
marimo.io/blog/newslet...
Doing a slightly better documentation than just the README.md
alonsosilvaallende.github.io/litelines/ge...
Feeback?
alonsosilvaallende.github.io/litelines/ge...
Feeback?
September 27, 2025 at 12:05 PM
Doing a slightly better documentation than just the README.md
alonsosilvaallende.github.io/litelines/ge...
Feeback?
alonsosilvaallende.github.io/litelines/ge...
Feeback?
Want to visualize the response format constraints on the LLM when working in a Jupyter notebook?
Then you might be interested in my new project `litelines`.
Litelines lets you visualize the selected path by the LLM.
It supports a Pydantic schema as a response format, as well as regular expressions.
Then you might be interested in my new project `litelines`.
Litelines lets you visualize the selected path by the LLM.
It supports a Pydantic schema as a response format, as well as regular expressions.
September 16, 2025 at 7:20 AM
Want to visualize the response format constraints on the LLM when working in a Jupyter notebook?
Then you might be interested in my new project `litelines`.
Litelines lets you visualize the selected path by the LLM.
It supports a Pydantic schema as a response format, as well as regular expressions.
Then you might be interested in my new project `litelines`.
Litelines lets you visualize the selected path by the LLM.
It supports a Pydantic schema as a response format, as well as regular expressions.
Here is the Hugging Face space to play with it: huggingface.co/spaces/alons...
Here is a Google Colab with the same information as the post:
nbviewer.org/gist/alonsos...
Here is a Google Colab with the same information as the post:
nbviewer.org/gist/alonsos...
July 23, 2025 at 7:19 PM
Here is the Hugging Face space to play with it: huggingface.co/spaces/alons...
Here is a Google Colab with the same information as the post:
nbviewer.org/gist/alonsos...
Here is a Google Colab with the same information as the post:
nbviewer.org/gist/alonsos...
New blog post: Constrain a language model not to use the letter 'e'
alonsosilvaallende.github.io/blog/posts/2...
In this post, I constrain a small language model (0.6B parameters) with a logits processor to accomplish something GPT-4o fails to do (see chatgpt.com/share/687908...).
alonsosilvaallende.github.io/blog/posts/2...
In this post, I constrain a small language model (0.6B parameters) with a logits processor to accomplish something GPT-4o fails to do (see chatgpt.com/share/687908...).
July 23, 2025 at 7:19 PM
New blog post: Constrain a language model not to use the letter 'e'
alonsosilvaallende.github.io/blog/posts/2...
In this post, I constrain a small language model (0.6B parameters) with a logits processor to accomplish something GPT-4o fails to do (see chatgpt.com/share/687908...).
alonsosilvaallende.github.io/blog/posts/2...
In this post, I constrain a small language model (0.6B parameters) with a logits processor to accomplish something GPT-4o fails to do (see chatgpt.com/share/687908...).
TIL how to add notebook cells running on @pyodide.org to Quarto projects, such as my blog:
alonsosilvaallende.github.io/til/posts/20...
It's very easy to change the Pyodide version if needed. Thanks @coatless.bsky.social for this great Quarto extension
alonsosilvaallende.github.io/til/posts/20...
It's very easy to change the Pyodide version if needed. Thanks @coatless.bsky.social for this great Quarto extension
July 18, 2025 at 3:35 PM
TIL how to add notebook cells running on @pyodide.org to Quarto projects, such as my blog:
alonsosilvaallende.github.io/til/posts/20...
It's very easy to change the Pyodide version if needed. Thanks @coatless.bsky.social for this great Quarto extension
alonsosilvaallende.github.io/til/posts/20...
It's very easy to change the Pyodide version if needed. Thanks @coatless.bsky.social for this great Quarto extension
New blog post: Understanding Logits Processors
alonsosilvaallende.github.io/blog/posts/2...
I start with a basic min length example, then force the model to check its answer, followed by making reasoning models stop thinking once they reach a thinking budget & finally force the model to think longer
alonsosilvaallende.github.io/blog/posts/2...
I start with a basic min length example, then force the model to check its answer, followed by making reasoning models stop thinking once they reach a thinking budget & finally force the model to think longer
July 16, 2025 at 3:31 PM
New blog post: Understanding Logits Processors
alonsosilvaallende.github.io/blog/posts/2...
I start with a basic min length example, then force the model to check its answer, followed by making reasoning models stop thinking once they reach a thinking budget & finally force the model to think longer
alonsosilvaallende.github.io/blog/posts/2...
I start with a basic min length example, then force the model to check its answer, followed by making reasoning models stop thinking once they reach a thinking budget & finally force the model to think longer
New blog post: Understanding Structured Outputs
alonsosilvaallende.github.io/blog/posts/2...
This is the power behind structured ouputs libraries such as Instructor and Marvin.
I provide a basic example of extraction, a slightly more complex one, then classification, and finally SO in WebAssembly.
alonsosilvaallende.github.io/blog/posts/2...
This is the power behind structured ouputs libraries such as Instructor and Marvin.
I provide a basic example of extraction, a slightly more complex one, then classification, and finally SO in WebAssembly.
July 11, 2025 at 12:21 PM
New blog post: Understanding Structured Outputs
alonsosilvaallende.github.io/blog/posts/2...
This is the power behind structured ouputs libraries such as Instructor and Marvin.
I provide a basic example of extraction, a slightly more complex one, then classification, and finally SO in WebAssembly.
alonsosilvaallende.github.io/blog/posts/2...
This is the power behind structured ouputs libraries such as Instructor and Marvin.
I provide a basic example of extraction, a slightly more complex one, then classification, and finally SO in WebAssembly.
New blog post: Understanding Function Calling
I provide a basic example of FC, then a slightly more complex example by allowing an LM to use Python. I explain the conversational response as a tool trick. Finally, FC in the browser by using WebAssembly
alonsosilvaallende.github.io/blog/posts/2...
I provide a basic example of FC, then a slightly more complex example by allowing an LM to use Python. I explain the conversational response as a tool trick. Finally, FC in the browser by using WebAssembly
alonsosilvaallende.github.io/blog/posts/2...
July 5, 2025 at 8:25 PM
New blog post: Understanding Function Calling
I provide a basic example of FC, then a slightly more complex example by allowing an LM to use Python. I explain the conversational response as a tool trick. Finally, FC in the browser by using WebAssembly
alonsosilvaallende.github.io/blog/posts/2...
I provide a basic example of FC, then a slightly more complex example by allowing an LM to use Python. I explain the conversational response as a tool trick. Finally, FC in the browser by using WebAssembly
alonsosilvaallende.github.io/blog/posts/2...
So I appear in the Celebration of 100 years of Bell Labs video with our LLM robots (a.k.a. IndustrialGPT) for less than a second :-D
www.youtube.com/watch?v=Fu_I...
www.youtube.com/watch?v=Fu_I...
June 30, 2025 at 9:59 AM
So I appear in the Celebration of 100 years of Bell Labs video with our LLM robots (a.k.a. IndustrialGPT) for less than a second :-D
www.youtube.com/watch?v=Fu_I...
www.youtube.com/watch?v=Fu_I...
New blog post: Understanding LLM Memory
alonsosilvaallende.github.io/blog/posts/2...
Using the Marimo extension for Quarto.
alonsosilvaallende.github.io/blog/posts/2...
Using the Marimo extension for Quarto.
June 28, 2025 at 10:44 PM
New blog post: Understanding LLM Memory
alonsosilvaallende.github.io/blog/posts/2...
Using the Marimo extension for Quarto.
alonsosilvaallende.github.io/blog/posts/2...
Using the Marimo extension for Quarto.
Celebrating Bell Labs’ 100 years anniversary #NokiaBellLabs100
June 17, 2025 at 9:40 AM
Celebrating Bell Labs’ 100 years anniversary #NokiaBellLabs100
TIL to do quizzes with JupyterQuiz:
alonsosilvaallende.github.io/til/posts/20...
I like this simple way to provide feedback to students. I’m using these quizzes to motivate my kid to learn new things.
alonsosilvaallende.github.io/til/posts/20...
I like this simple way to provide feedback to students. I’m using these quizzes to motivate my kid to learn new things.
June 16, 2025 at 7:39 PM
TIL to do quizzes with JupyterQuiz:
alonsosilvaallende.github.io/til/posts/20...
I like this simple way to provide feedback to students. I’m using these quizzes to motivate my kid to learn new things.
alonsosilvaallende.github.io/til/posts/20...
I like this simple way to provide feedback to students. I’m using these quizzes to motivate my kid to learn new things.
Finally meeting @remilouf.bsky.social !
May 20, 2025 at 5:59 PM
Finally meeting @remilouf.bsky.social !
More coverage of our work at @ieeespectrum.bsky.social 🤩
April 23, 2025 at 7:43 AM
More coverage of our work at @ieeespectrum.bsky.social 🤩
Chat Templates
Chat is an abstraction. LLMs know nothing about system messages, user messages, etc. The translation between messages and the received string representation is handled by the chat template.
Code: github.com/alonsosilvaa...
To try it:
uv run raw.githubusercontent.com/alonsosilvaa...
Chat is an abstraction. LLMs know nothing about system messages, user messages, etc. The translation between messages and the received string representation is handled by the chat template.
Code: github.com/alonsosilvaa...
To try it:
uv run raw.githubusercontent.com/alonsosilvaa...
April 20, 2025 at 10:04 PM
Chat Templates
Chat is an abstraction. LLMs know nothing about system messages, user messages, etc. The translation between messages and the received string representation is handled by the chat template.
Code: github.com/alonsosilvaa...
To try it:
uv run raw.githubusercontent.com/alonsosilvaa...
Chat is an abstraction. LLMs know nothing about system messages, user messages, etc. The translation between messages and the received string representation is handled by the chat template.
Code: github.com/alonsosilvaa...
To try it:
uv run raw.githubusercontent.com/alonsosilvaa...
Next token
The code can be found here: github.com/alonsosilvaa...
If you have uv installed, you can try it with this command:
uv run raw.githubusercontent.com/alonsosilvaa...
The code can be found here: github.com/alonsosilvaa...
If you have uv installed, you can try it with this command:
uv run raw.githubusercontent.com/alonsosilvaa...
April 19, 2025 at 11:31 PM
Next token
The code can be found here: github.com/alonsosilvaa...
If you have uv installed, you can try it with this command:
uv run raw.githubusercontent.com/alonsosilvaa...
The code can be found here: github.com/alonsosilvaa...
If you have uv installed, you can try it with this command:
uv run raw.githubusercontent.com/alonsosilvaa...
The answer is wrong but DeepSeek's reasoning traces are always fun to watch :-)
February 16, 2025 at 9:43 PM
The answer is wrong but DeepSeek's reasoning traces are always fun to watch :-)
I use Kùzu but I must admit these are cute 😜
January 16, 2025 at 9:10 PM
I use Kùzu but I must admit these are cute 😜
My 2025 resolution is to take more risks
January 1, 2025 at 10:10 PM
My 2025 resolution is to take more risks
The pictures from the @iadevs.bsky.social conference are now available:
photos.google.com/share/AF1Qip...
It was a super cool conference!
photos.google.com/share/AF1Qip...
It was a super cool conference!
December 16, 2024 at 9:01 PM
The pictures from the @iadevs.bsky.social conference are now available:
photos.google.com/share/AF1Qip...
It was a super cool conference!
photos.google.com/share/AF1Qip...
It was a super cool conference!
Mañana presento “Creación de agentes basados en gráficos de conocimiento con generación de texto estructurado y modelos open-weights” en el Centro de Modelamiento Matemático de la Universidad de Chile
www.cmm.uchile.cl?p=59116
www.cmm.uchile.cl?p=59116
December 12, 2024 at 10:19 PM
Mañana presento “Creación de agentes basados en gráficos de conocimiento con generación de texto estructurado y modelos open-weights” en el Centro de Modelamiento Matemático de la Universidad de Chile
www.cmm.uchile.cl?p=59116
www.cmm.uchile.cl?p=59116
Hoy estaré presentando Mejorando la generación aumentada con recuperación (RAG) con la búsqueda híbrida en la conferencia EVIC 2024:
www.evic.cl
¡Nos vemos!
www.evic.cl
¡Nos vemos!
December 12, 2024 at 8:28 AM
Hoy estaré presentando Mejorando la generación aumentada con recuperación (RAG) con la búsqueda híbrida en la conferencia EVIC 2024:
www.evic.cl
¡Nos vemos!
www.evic.cl
¡Nos vemos!