Whole-Brain Fly Simulation Hits 91% Behavior Accuracy With No Training
alexcovo_eth · x · 2026-09-12
A widely shared result on whole fruit fly brain simulation: instead of training on massive data, researchers ported the biological brain's physical circuitry directly into a simulator. The digital fly achieved 91% behavior accuracy using only a complete connectome, synapse-count-based weights, excitatory/inhibitory neuron classification, and simple LIF neuron models. The result suggests brain topology itself may encode much of behavior—evidence for the whole-brain emulation path, though scaling from fly-level to the human brain's 86 billion neurons remains an open question.
More from Research
- WISD: Distill Training Signal from Weaker Quantized Copies of Reasoning Models — jet-ai · 2026-10-31
- LLM Acknowledgments in Bio Preprints Hit 17% in Three Years, Near Specialized AI Tools — teortaxesTex · 2026-10-31
- StudentBench author: AI tutors boosting human intelligence is the use-case to hill-climb — gharik · 2026-10-31
- Spatial Hub aggregates ~5000 public COLMAP-compatible spatial datasets for world models — janusch_patas · 2026-10-31
- ARC Prize 2026 closes in 3 days as Yi-Chia Chen tops ARC-AGI-3 at 75.56% — GregKamradt · 2026-10-31
- NVIDIA Nemotron 3.5 scales LLM self-verification, +9-10 points on agentic benchmarks — sarahcat21 · 2026-10-31