Axel Højmark
Research Scientist, Apollo Research · London
About
I work on AI safety and alignment.
I'm a research scientist at Apollo Research, studying ways to detect and mitigate deception in frontier models.
Previously:
- A MATS scholar (Summer 2024) with Marius Hobbhahn and Jérémy Scheurer, researching scaling laws for LM agents.
- I contributed long-horizon agent tasks to METR's catastrophic-risk evaluation suite through an open bounty.
- BSc in Machine Learning and Data Science from the University of Copenhagen. Started an MSc but left to work on AI safety full-time.
Outside of work, I'm very interested in longevity, economics, and human intelligence enhancement.
Highlighted research
-
Stress Testing Deliberative Alignment for Anti-Scheming Training
Stress-tests an anti-scheming training intervention across 26 out-of-distribution evaluations. Covert-action rates drop sharply but are not eliminated, and we find causal evidence that rising evaluation awareness partly drives the reduction.
-
Forecasting Frontier Language Model Agent Capabilities
Forecasts frontier agent capabilities on SWE-Bench Verified, Cybench, and RE-Bench through 2026, using a Release Date → Elo → Benchmark pipeline validated by backtesting six methods on open-source LLMs.