Research map / Probabilistic and scientific modeling / Uncertainty and Gaussian processes
Autoinverse: Uncertainty Aware Inversion of Neural Networks
Navid Ansari, Hans-peter Seidel, Nima Vahidi Ferdowsi, Vahid Babaei
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 |
|---|---|---|---|
| Add uncertainty-aware term (UANA) to NA | NA · multi-joint robot | NFP error: 3.24 \pm 0.51 -> 3.21 \pm 1.48 | Improved |
| Add uncertainty-aware term to tandem (UA-tandem) | tandem · multi-joint robot | NFP error: 4.42 \pm 1.56 -> 4.04 \pm 0.38 | Improved |
| Add uncertainty-aware term (UANA) to NA | NA · multi-joint robot | surrogate error: 1.99 \pm 0.05 -> 9.13 \pm 6.08 | Hurt |
| Use ensemble of networks in NA (NA ensemble) without uncertainty term (ablation) | NA · multi-joint robot | NFP error: 1.17 \pm 0.32 | Mixed |
| Use only ReLU activations in the ensemble instead of diverse activations (ablation) | UANA · spectral printer, sparse LC data | count of LC>=0.4 solutions: 0 \pm 0 -> 19 \pm 4 | Hurt |
| Use UANA instead of NA on data with sparse LC channel | NA · spectral printer, sparse data | count of LC>=0.4 solutions: 1604 \pm 11 -> 0 \pm 0 | Improved |
| Use UANA instead of NA on data with noisy LC channel | NA · spectral printer, noisy data | count of LC>=0.4 solutions: 1029 \pm 23 -> 0 \pm 0 | Improved |
| Use UA-tandem instead of tandem on sparse LC data | tandem · spectral printer, sparse data | count of LC>=0.4 solutions: 52 \pm 30 -> 0 \pm 0 | Improved |
| Use UANA instead of NA on NFP error with standard data | NA · spectral printer, standard data | NFP error: 1.57 \pm 0.001 -> 6.30 \pm 0.031 | Improved |
| Add boundary loss regularization to NA | NA · spectral printer | ink density distribution | Mixed |
| Add smoothness regularization to NA | NA · soft robot | design plausibility | Improved |
| Use UANA without any regularization instead of regularized NA | NA with smoothness term · soft robot | design plausibility | No clear effect |
Closest papers by meaning
- Uncertainty Quantification for Forward and Inverse Problems of PDEs via Latent Global Evolution (AAAI 2024 · similarity 0.84)
- Network Inversion of Convolutional Neural Nets (Student Abstract) (AAAI 2025 · similarity 0.84)
- Image Reconstruction Via Autoencoding Sequential Deep Image Prior (NeurIPS 2024 · similarity 0.83)
- A Study of Bayesian Neural Network Surrogates for Bayesian Optimization (ICLR 2024 · similarity 0.83)
- Auto-Compressing Networks (NeurIPS 2025 · similarity 0.83)
- BARNN: A Bayesian Autoregressive and Recurrent Neural Network (ICML 2025 · 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.
