Lukas Paul Achatius Galke Poech

Teaching Portfolio !!Link opens in a new tab

Vis Scopus-profil

Telefon+4565509526
E-mailgalkeimada.sdudk

Forskningsområder

My research advances AI Safety by investigating how and why language models behave they way they do.

Emneord

Maskinlæring
AI Safety
Natural Language Processing

Fingeraftryk

Dyk ned i forskningsemnerne, hvor Lukas Paul Achatius Galke Poech er aktiv. Disse emneordskoncepter kommer fra denne persons arbejde. Sammen danner de et unikt fingerprint.

1 Lignende profiler

30 Konferencebidrag i proceedings
7 Tidsskriftartikel
3 Rapport
3 Andet bidrag
Mere

Deep neural networks and humans both benefit from compositional language structure
Galke, L., Ram, Y. & Raviv, L., 30. dec. 2024, I: Nature Communications. 15, 1, 13 s., 10816.
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › peer review

Åben adgang
Fil
10 Downloads (Pure)
Learning and communication pressures in neural networks: Lessons from emergent communication
Galke, L. P. A. & Raviv, L., 2025, I: Language Development Research. 5, 1, s. 116-143
Publikation: Bidrag til tidsskrift › Tidsskriftartikel › Forskning › peer review

Åben adgang
Fil
102 Downloads (Pure)
Isolating Culture Neurons in Multilingual Large Language Models
Namazifard, D. & Galke, L., 2025, Findings of the Association for Computational Linguistics: IJCNLP-AACL 2025. Association for Computational Linguistics, s. 768–785
Publikation: Kapitel i bog/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review
Chain of Summaries: Summarization Through Iterative Questioning
Brach, W. & Poech, L. G., 2026, 10 s.
Publikation: Andet › Andet bidrag › Forskning

Fil
1 Downloads (Pure)
DeToNATION: Decoupled Torch Network-Aware Training on Interlinked Online Nodes
From, M. H., Nielsen, J., Poech, L. G. & Schneider-Kamp, P., 2026, Proceedings of the AAAI Conference on Artificial Intelligence: AAAI-26. AAAI Press, Bind 40. s. 21128-21135 8 s.
Publikation: Kapitel i bog/rapport/konference-proceeding › Konferencebidrag i proceedings › Forskning › peer review

Åben adgang
Fil
3 Downloads (Pure)

Lukas Paul Achatius Galke Poech

Personlig profil

Forskningsområder

Emneord

Fingeraftryk

Samarbejde og topforskningsområder i de sidste fem år

Dyk ned i detaljerne

Deep neural networks and humans both benefit from compositional language structure

Learning and communication pressures in neural networks: Lessons from emergent communication

Isolating Culture Neurons in Multilingual Large Language Models

Chain of Summaries: Summarization Through Iterative Questioning

DeToNATION: Decoupled Torch Network-Aware Training on Interlinked Online Nodes

Lukas Paul Achatius Galke Poech

Personlig profil

Forskningsområder

Emneord

Fingeraftryk

Samarbejde og topforskningsområder i de sidste fem år

Publikationer

Deep neural networks and humans both benefit from compositional language structure

Learning and communication pressures in neural networks: Lessons from emergent communication

Isolating Culture Neurons in Multilingual Large Language Models

Chain of Summaries: Summarization Through Iterative Questioning

DeToNATION: Decoupled Torch Network-Aware Training on Interlinked Online Nodes