Rodrigo Kazu Siqueira
Frontier Llm Evaluation Contractor Ai Benchmark Problem Author @Alignerr
Signup · Get unlimited contacts
WORK HISTORY
Frontier Llm Evaluation Contractor Ai Benchmark Problem Author @Alignerr
I work as a Frontier LLM Evaluation Contractor at Alignerr, designing graduate-level expert problems in GPQA and SciCode/STEM benchmark formats to evaluate and train frontier language models. Each problem I produce must be genuinely unsolvable by current AI without deep domain expertise — requiring multi-step reasoning, method-level knowledge, and awareness of failure modes that only practitioners encounter. I author the complete package: the problem, the expert reference solution, weighted rubric criteria with inline ground truth, and unit tests where applicable. These rubrics function directly as reward signals for reinforcement learning, meaning the quality and difficulty calibration of my work shapes how frontier models are trained in life sciences and computational biology.
EDUCATION
Federal University of Rio de Janeiro
Master's degree in Morphological Sciences (Focused on Neuroscience), Neurobiology and Neurosciences
University of Reading
Doctor of Philosophy (PhD), Cybernetics/Bioengineering
Federal University of Rio de Janeiro
Bachelor's degree, Zoology/Animal Biology
ABOUT RODRIGO KAZU SIQUEIRA
https://neuroscience.cam.ac.uk/member/rodrigokazu/ORCID: Author ID: 56•••••••00 I am a Brazilian-born British Neuroscientist with a PhD in Cybernetics and currently focused on discovering novel cell-specific insights from single-cell multiome profiling of the motor cortex in ALS donors at SITraN.https://sheffield.ac.uk/smph/people/academic/neuroscience/rodrigo-kazu-siqueira I also have a focus in computational neuroscience and complex networks at the Wellcome Sanger Institute and Cambridge Neuroscience (University of Cambridge). Previously, I helped people walk again by modelling the tactile responses of the foot sole aiming at real-time feedback prosthetics and won the INSIGNEO Showcase prize for best research with this study in 2019 and the breakthrough was recently published in the journal research interests include ALS, spatial transcriptomics, bioinformatics, general intelligence, network neuroscience, brain development and brain evolution. Additionally, I am trying to understand cell fates in tumour associated macrophages in Glioblastoma Multiforme.Most of my code can be accessed at https://github.com/rodrigokazu;Google Scholar: & hl=enI am also a Sheffield Data Science and AI community member.
This profile is compiled from publicly available professional sources. Unifers is not affiliated with or endorsed by LinkedIn. Request removal of this profile.