Our new book: Agentic Large Language Models
Published:
After a year and a half of writing, our textbook on agentic LLMs is out as a preprint, with a Springer edition and a new master course to go with it.
Niki van Stein
Associate Professor in Explainable AI for ML and Optimization
- Institute
- LIACS
- Based in
- Leiden
Posts by Tags
agents
automatic-algorithm-design
From configurations to code: evolving AutoML pipelines with LLaMEA
Published:
LACE lets a language model evolve complete, readable machine learning pipelines, the topic of my keynote at the AutoML Conference 2026.
Looking inside the model: mechanistic interpretability, and why LLaMEA needs it
Published:
A plain-language tour of our paper “Every Component Is a Lookup”, and why understanding how LLMs make design decisions matters for evolving code.
LLaMEA: Combining the Power of Evolution and Large Language Models
Published:
LLaMEA, a pioneering framework combining Evolution and Large Language Models.
automl
From configurations to code: evolving AutoML pipelines with LLaMEA
Published:
LACE lets a language model evolve complete, readable machine learning pipelines, the topic of my keynote at the AutoML Conference 2026.
bias
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
book
Our new book: Agentic Large Language Models
Published:
After a year and a half of writing, our textbook on agentic LLMs is out as a preprint, with a Springer edition and a new master course to go with it.
conference
Gecco 2023 - DoE2Vec
Published:
Gecco 2023
Gecco 2023 - Deep Bias
Published:
Gecco 2023
deep-learning
LLaMEA: Combining the Power of Evolution and Large Language Models
Published:
LLaMEA, a pioneering framework combining Evolution and Large Language Models.
Reflecting on CIMPLO
Published:
A Journey Through Predictive Maintenance
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
evolutionary-computation
From configurations to code: evolving AutoML pipelines with LLaMEA
Published:
LACE lets a language model evolve complete, readable machine learning pipelines, the topic of my keynote at the AutoML Conference 2026.
explainable-AI
Reflecting on CIMPLO
Published:
A Journey Through Predictive Maintenance
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
explainable-ai
Looking inside the model: mechanistic interpretability, and why LLaMEA needs it
Published:
A plain-language tour of our paper “Every Component Is a Lookup”, and why understanding how LLMs make design decisions matters for evolving code.
gecco
Gecco 2023 - DoE2Vec
Published:
Gecco 2023
Gecco 2023 - Deep Bias
Published:
Gecco 2023
heuristics
LLaMEA: Combining the Power of Evolution and Large Language Models
Published:
LLaMEA, a pioneering framework combining Evolution and Large Language Models.
Reflecting on CIMPLO
Published:
A Journey Through Predictive Maintenance
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
llamea
From configurations to code: evolving AutoML pipelines with LLaMEA
Published:
LACE lets a language model evolve complete, readable machine learning pipelines, the topic of my keynote at the AutoML Conference 2026.
Looking inside the model: mechanistic interpretability, and why LLaMEA needs it
Published:
A plain-language tour of our paper “Every Component Is a Lookup”, and why understanding how LLMs make design decisions matters for evolving code.
llm
Our new book: Agentic Large Language Models
Published:
After a year and a half of writing, our textbook on agentic LLMs is out as a preprint, with a Springer edition and a new master course to go with it.
From configurations to code: evolving AutoML pipelines with LLaMEA
Published:
LACE lets a language model evolve complete, readable machine learning pipelines, the topic of my keynote at the AutoML Conference 2026.
Looking inside the model: mechanistic interpretability, and why LLaMEA needs it
Published:
A plain-language tour of our paper “Every Component Is a Lookup”, and why understanding how LLMs make design decisions matters for evolving code.
LLaMEA: Combining the Power of Evolution and Large Language Models
Published:
LLaMEA, a pioneering framework combining Evolution and Large Language Models.
mechanistic-interpretability
Looking inside the model: mechanistic interpretability, and why LLaMEA needs it
Published:
A plain-language tour of our paper “Every Component Is a Lookup”, and why understanding how LLMs make design decisions matters for evolving code.
optimization
LLaMEA: Combining the Power of Evolution and Large Language Models
Published:
LLaMEA, a pioneering framework combining Evolution and Large Language Models.
Reflecting on CIMPLO
Published:
A Journey Through Predictive Maintenance
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
presentation
Gecco 2023 - DoE2Vec
Published:
Gecco 2023
Gecco 2023 - Deep Bias
Published:
Gecco 2023
structural-bias
Structural Bias, What is it and why do we care?
Published:
An introduction to Structural Bias in search heuristics.
teaching
Our new book: Agentic Large Language Models
Published:
After a year and a half of writing, our textbook on agentic LLMs is out as a preprint, with a Springer edition and a new master course to go with it.