Research map / Learning theory and algorithms / Prediction and decision losses
Adaptive stimulus selection for optimizing neural population responses
Benjamin Cowley, Ryan Williamson, Katerina Clemens, Matthew Smith, Byron M. Yu
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use Adept instead of random image selection | Adept · CNN testbed | mean neuron response: 15.5% larger | Improved |
| use Adept instead of random image selection | Adept · CNN testbed | response scatter (PCA eigenvalue ratio) | Improved |
| population objective (Adept) instead of single-neuron objectives (Adept-1, Adept-50, genetic-50) (ablation) | Adept · CNN testbed | mean response and eigenvalue ratio | Improved |
| keep only norm term in objective (Adept-norm) (ablation) | Adept · CNN testbed | mean response: p = 0.65 | No clear effect |
| keep only norm term in objective (Adept-norm) (ablation) | Adept · CNN testbed | response scatter (eigenvalue ratio): p < 0.005 | Hurt |
| keep only average-distance term (Adept-avgdist) (ablation) | Adept · CNN testbed | mean response: p < 10 | Hurt |
| keep only average-distance term (Adept-avgdist) (ablation) | Adept · CNN testbed | response scatter (eigenvalue ratio): p = 0.62 | No clear effect |
| use embedding layer other than middle layer of GoogLeNet (ablation) | Adept feature embeddings · CNN testbed | correlation predicted vs actual objective: 0.6 | No clear effect |
| use Adept instead of random selection in macaque V4 | Adept · V4 recordings, 7 sessions | mean response: 19.5% higher | Improved |
| use Adept in V4 for response scatter | Adept · V4 recordings | eigenvalue ratio by PC dimension | Mixed |
| vary CNN layer used for V4 feature embeddings (deeper 5-10 vs 1-4) (ablation) | Adept feature embeddings · V4 recordings | correlation predicted vs actual objective | Improved |
| replace CNN embeddings with V4 responses from another day (ablation) | Adept feature embeddings · V4 recordings | correlation predicted vs actual objective: 0.25 -> 0.5 | Improved |
Closest papers by meaning
- Optimal Adaptive Electrode Selection to Maximize Simultaneously Recorded Neuron Yield (NeurIPS 2020 · similarity 0.87)
- NEvo: Neural-Guided Evolutionary Video Synthesis for Dynamic Visual Selectivity (NeurIPS 2026 · similarity 0.86)
- Information-based Adaptive Stimulus Selection to Optimize Communication Efficiency in Brain-Computer Interfaces (NeurIPS 2018 · similarity 0.85)
- Modeling Dynamic Neural Activity by combining Naturalistic Video Stimuli and Stimulus-independent Latent Factors (NeurIPS 2025 · similarity 0.84)
- Benchmarking Out-of-Distribution Generalization Capabilities of DNN-based Encoding Models for the Ventral Visual Cortex. (NeurIPS 2024 · similarity 0.83)
- Retrospective for the Dynamic Sensorium Competition for predicting large-scale mouse primary visual cortex activity from videos (NeurIPS 2024 · similarity 0.83)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
