If you like this research, contact us to do a hard and cool bachelor or master thesis in the group, see list of potential topics.
Follow our work on: Google Scholar, RSS, LinkedIn.
If you like this research, contact us to do a hard and cool bachelor or master thesis in the group, see list of potential topics.
Follow our work on: Google Scholar, RSS, LinkedIn.
awesome open list of pointers about open science for software and computational science
Chaos engineering systems invented at KTH Royal Institute of Technology. http://urn.kb.se/resolve?urn=urn:nbn:se:kth:diva-320638
Empirical experiments on agentic safety: does model-baked-in refusal or harness-level filtering stop catastrophic tool calls?
Automatic Repair Framework with LLMs ❤️ https://arxiv.org/pdf/2409.18952
Claude Code (Sonnet 5) on all 854 active bugs of Defects4J 3.0.1: 851 test-adequate patches (99.6%), with harness, session logs and contamination audit