Contact us

Aleph Alpha
Blog

Research, engineering and perspectives from the Aleph Alpha team.

Mind the Tokenizer: Why a Single Space Decides What Multiple-Choice Evaluations Measure

Research 20/10/2026
Balancing Experts at Two Scales: Exact Quantile Balancing and Load-Error Injection

Research 14/10/2026
PriorBPB: Reducing the Answer Length Confound of Bits-per-Byte While Improving Accuracy Prediction

Research 13/10/2026
Specialised LLMs punch above their weight

Research 08/10/2026
Designing Kolibri: Architecture Trade-Offs from First Principles

Research 07/10/2026
Generalization or Memorization? What Contaminated Training Data Does to Benchmark Scores

Research 06/10/2026
Kolibri Has Landed: A Sovereign Open-Weight Model

Research 03/10/2026
Scaling Pre-Training in Practice: A Hierarchical Approach

Research 30/09/2026
Training on the Party Line: Chinese Political Influence on LLMs in China and the World

Research 28/09/2026
Through the Valley of Tears: Cold Starting German Reasoning in LLMs

Research 24/09/2026