Elamur Research⌕
☰
home / authors / Matthew Rahtz

Matthew Rahtz

Matthew Rahtz has 3 papers in our library, in 2023–2024, most of them on Language models & pretraining, Evaluation & benchmarks, Interpretability; often with Janos Kramar, Tom Lieberum, Vladimir Mikulik.

3 papers in our library · 2023–2024

OpenAlex · h-index 7 · 14 works · 2,335 citations OpenAlex, 2026-10-14

📄 3 works in our library · 🔗 Cited 3,507 times in our library · 1 more in our library without a page yet · ⭐ Paper Stars ranking: #7554

The profile is the library's match of the name; two people with one name are not always told apart.

Elsewhere: Google Scholar · Semantic Scholar · dblp · ACL Anthology

 Affiliations

🎓ETH ZurichZurich · Switzerland2017–2019 · 2 papers

From the bylines of their papers (OpenAlex, 2026-10-14), ORCID and Wikidata; institutions on papers are matched by machine. 2 more institutions on single bylines are not listed.

 Fields, topics and concepts

Of its works in our library, with the number of works on each.

📚 Shelves 8The field of science this work belongs to (OpenAlex fields).Artificial Intelligence · 3 worksComputer Science · 3 worksArts and Humanities · 2 worksLanguage and Linguistics · 2 worksEngineering · 1 workSafety Research · 1 workSafety, Risk, Reliability and Quality · 1 workSocial Sciences · 1 work

🧭 Topics 3Research subjects and languages the work is about.Interpretability · 1 workLanguage models & pretraining · 1 workEvaluation & benchmarks · 1 work

💡 Concepts 1Key ideas and methods explained on our concept pages.Scaling laws · 1 work

 Works by year

2020s

20241 paper20232 papers

 Papers

The most cited first (OpenAlex and our library).

  1. Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in Chinchilla (2023 · 58 citations)
    Circuit analysis has mostly been demonstrated on small networks, such as two-layer attention-only transformers or the 117M GPT-2.
  2. The Hydra Effect: Emergent Self-repair in Language Model Computations (2023 · 49 citations)
    The paper asks how far ablation can be trusted for working out which parts of a large language model drive its predictions.
  3. Evaluating Frontier Models for Dangerous Capabilities (2024 · 36 citations)
    Popular benchmarks measure broad skills such as coding or maths, and most safety tests check whether a model refuses harmful requests.

All works

4 works of theirs in our library, the most cited first; 3 with a page here, the others link their original until we summarise them.

  1. Gemma 2: Improving Open Language Models at a Practical Size ↗arXiv:2408.00118 · no summary yetGemma Team et al.2024 · 1,302 citations
  2. Does Circuit Analysis Interpretability Scale? Evidence from Multiple Choice Capabilities in ChinchillaarXiv:2307.09458Tom Lieberum, Matthew Rahtz, Janos Kramar, et al.2023 · 58 citations
  3. The Hydra Effect: Emergent Self-repair in Language Model ComputationsarXiv:2307.15771Thomas McGrath, Matthew Rahtz, Janos Kramar, et al.2023 · 49 citations
  4. Evaluating Frontier Models for Dangerous CapabilitiesarXiv:2403.13793Mary Phuong, Matthew Aitchison, Elliot Catt, et al.2024 · 36 citations

 Co-authors here

12 co-authors with a page, the most papers together first.

 Cited by

Papers of our library cite their works 3,507 times (of the 149,439 papers whose bibliographies we read); 897 citing papers with a page here →.

Updated

This page in Markdown → https://md.research.elamur.ai/free/en/authors/matthew-rahtz/