PathNet: Evolution Channels Gradient Descent in Super Neural Networks (2017 · 667 citations)
The paper asks how one very large neural network could learn many tasks one after another. The goal is to reuse earlier knowledge without overwriting it, which is known as catastrophic forgetting. The authors propose PathNet, a modular network of several layers, each with many small modules. A population of…
Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution (2023 · 174 citations)
Hand-written prompting strategies like Chain-of-Thought help LLMs reason, but they are generic and seldom the best choice for a particular task.

All works

3 works of theirs in our library, the most cited first; 2 with a page here, the others link their original until we summarise them.

PathNet: Evolution Channels Gradient Descent in Super Neural NetworksarXiv:1701.08734Chrisantha Fernando, Dylan Banarse, Charles Blundell, et al.2017 · 667 citations
Promptbreeder: Self-Referential Self-Improvement Via Prompt EvolutionarXiv:2309.16797Chrisantha Fernando, Dylan Banarse, Henryk Michalewski, et al.2023 · 174 citations
Perception Test: A Diagnostic Benchmark for Multimodal Video Models ↗neural information processing systems · no summary yetViorica Pătrăucean et al.2023 · 113 citations

Co-authors here

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

Where their works here were published, the most works first, with the editions of those that have a page.

Papers of our library cite their works 2,533 times (of the 149,439 papers whose bibliographies we read); 615 citing papers with a page here →.

Updated 2 November 2026