Virtual animal models: useful for enumeration, risky for claiming real mechanisms
neuroecology · x · 2026-11-05
@neuroecology wraps up the virtual-animals-for-neuroscience thread: he's suspicious of taking virtual systems and claiming "this is what animal X is doing" (his own papers notwithstanding), but finds it interesting to enumerate how similar animals could be built—as a way of mapping the space of possible neural mechanisms.
Related event: Neuroscientists Bet on Deep RL "Virtual Animals" for Brain Research(5 posts)→
More from Research
- For early-career scientists, the cost of AI tools becomes a career problem — AnnaCiaunica · 2026-11-05
- Hyper-Connections debate: mHC slightly beats iHC, xHC gains unclear — YouJiacheng · 2026-11-05
- Biomedical benchmark: prewritten Cypher MCP tools hit 98% vs 85% for Text2Cypher — JeremyCMorgan · 2026-11-05
- Nobel laureate David Baker lectures at Royal Society and Oxford on de novo enzyme design — mmbronstein · 2026-11-05
- Benchmarks need Kaggle-style public/private sets to stop data leakage — HanchungLee · 2026-11-05
- Three distinct reasons researchers study looped transformers, explained — Kangwook_Lee · 2026-11-05