The project is going to carry out mechanistic analysis of multimodal and reasoning large language models (LLMs). The project extends our previous work, which has been carried out in a set of small NAISS projects. The overall goal is to develop new methods to discover computation mechanisms inside of pre-trained LLMs. The main goals of the current project is to extend our previous results to new types of multimodal LLMs (mainly text/audio and text/image models at this point), and in particular we are going to investigate reasoning mechisms in these models. The practical work will mainly be carried out using public LLM benchmarks and open-weight LLMs. We are not planning to work with any text containing sensitive information. The project will be carried out by members of the NLP@DSAI group at Chalmers, led by professor Richard Johansson (see the group’s page https://dsai-nlp.github.io/). The text/vision and reasoning work will also involve collaborators from Zenseact, while the audio-focused work is a collaboration with AI Sweden The practical work will mainly be carried out by PhD and MSc students in the NLP@DSAI group. The resulting research will be presented in a number of planned articles that will be submitted to the EMNLP conference, one of the two top conferences in the NLP field.