Research map / Label-efficient and robust learning / Self-supervised and contrastive
Towards robust and generalizable representations of extracellular data using contrastive learning
Ankit Vishnubhotla, Charlotte Loh, Akash Srivastava, Liam Paninski, Cole Lincoln Hurwitz
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 |
|---|---|---|---|
| replace denoised PCA with CEED features (11 channels) | spike re-sorting · 10 neuron ID | ARI: .39 -> .89 | Improved |
| replace denoised PCA with CEED features (5 channels) | spike re-sorting · 10 neuron ID | ARI: .46 -> .83 | Improved |
| replace denoised PCA with CEED (11 channels) | spike re-sorting · 400 neuron ID | ARI: .45 -> .79 | Improved |
| replace denoised PCA with CEED (11 channels) | spike re-sorting · 390 neuron OOD | ARI: .46 -> .78 | Improved |
| replace non-linear autoencoder with CEED (11 channels) | spike re-sorting · 390 neuron OOD | ARI: .28 -> .78 | Improved |
| replace autoencoder with CEED (5 channels) | spike re-sorting · 10 neuron ID | ARI: .43 -> .83 | Improved |
| use 11 channels instead of 5 for CEED (ablation) | CEED · 10 neuron ID | ARI: .83 -> .89 | Improved |
| use 11 channels instead of 5 for CEED (ablation) | CEED · 390 neuron OOD | ARI: .77 -> .78 | No clear effect |
| use 11 channels instead of 5 for PCA (ablation) | denoised PCA · 10 neuron ID | ARI: .46 -> .39 | Hurt |
| use CEED instead of WaveMap for zero-shot cell-type clustering | cell-type classification · monkey 625 templates | classifier accuracy: 88.8% -> 93.4% | Improved |
| compare CEED to the WaveMap accuracy reported with 8 clusters | cell-type classification · monkey 625 templates | classifier accuracy: 91% -> 93.4% | Improved |
| denoise waveforms with YASS before PCA (ablation) | PCA baseline · spike sorting | clustering quality | Improved |
Closest papers by meaning
- TRACE: Contrastive learning for multi-trial time series data in neuroscience (NeurIPS 2025 · similarity 0.86)
- Reproducibility of predictive networks for mouse visual cortex (NeurIPS 2024 · similarity 0.84)
- Contrastive Consolidation of Top-Down Modulations Achieves Sparsely Supervised Continual Learning (NeurIPS 2025 · similarity 0.84)
- Neuron Platonic Intrinsic Representation From Dynamics Using Contrastive Learning (ICLR 2025 · similarity 0.83)
- Learning Multimodal Volumetric Features for Large-Scale Neuron Tracing (AAAI 2024 · similarity 0.82)
- ActSort: An active-learning accelerated cell sorting algorithm for large-scale calcium imaging datasets (NeurIPS 2024 · similarity 0.82)
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.
