Sparse regression and feature extraction are the cornerstones of knowledge discovery from massive data. Their goal is to discover interpretable and predictive models that provide simple relationships among scientific variables. While the statistical tools for model discovery are well established in the context of linear regression, their generalization to nonlinear regression in material modeling is highly problem‐specific and insufficiently understood. Here we explore the potential of neural networks for automatic model discovery and induce sparsity by a hybrid approach that combines two strategies: regularization and physical constraints. We integrate the concept of L_p regularization for subset selection with constitutive neural networks that leverage our domain knowledge in kinematics and thermodynamics. We train our networks with both, synthetic and real data, and perform several thousand discovery runs to infer common guidelines and trends: L₂ regularization or ridge regression is unsuitable for model discovery; L₁ regularization or lasso promotes sparsity, but induces strong bias that may aggressively change the results; only L₀ regularization allows us to transparently fine‐tune the trade‐off between interpretability and predictability, simplicity and accuracy, and bias and variance. With these insights, we demonstrate that L_p regularized constitutive neural networks can simultaneously discover both, interpretable models and physically meaningful parameters. We anticipate that our findings will generalize to alternative discovery techniques such as sparse and symbolic regression, and to other domains such as biology, chemistry, or medicine. Our ability to automatically discover material models from data could have tremendous applications in generative material design and open new opportunities to manipulate matter, alter properties of existing materials, and discover new materials with user‐defined properties.

Eating less meat is associated with a healthier body and planet. Yet, we remain reluctant to switch to a plant-based diet, largely due to the sensory experience of plant-based meat. Food scientists characterize meat using a double compression test, which only probes one-dimensional behavior. Here we use tension, compression, and shear tests–combined with constitutive neural networks–to automatically discover the behavior of eight plant-based and animal meats across the entire three-dimensional spectrum. We find that plant-based sausage and hotdog, with stiffnesses of 95.9 ± 14.1 kPa and 38.7 ± 3.0 kPa, successfully mimic their animal counterparts, with 63.5 ± 45.7 kPa and 44.3 ± 13.2 kPa, while tofurky is twice as stiff, and tofu is twice as soft. Strikingly, a complementary food tasting survey produces in nearly identical stiffness rankings for all eight products (ρ = 0.833, p = 0.015). Probing the fully three-dimensional signature of meats is critical to understand subtle differences in texture that may result in a different perception of taste. Our data and code are freely available at https://github.com/LivingMatterLab/CANN

The texture of meat is one of the most important features to mimic when developing meat analogs. Both protein source and processing method impact the texture of the final product. We can distinguish three types of mechanical tests to quantify the textural differences between meat and meat analogs: puncture type, rheological torsion tests, and classical mechanical tests of tension, compression, and bending. Here, we compile the shear force and stiffness values of whole and comminuted meats and meat analogs from the two most popular tests for meat, the Warner–Bratzler shear test and the double-compression texture profile analysis. Our results suggest that, with the right fine-tuning, today’s meat analogs are well capable of mimicking the mechanics of real meat. While Warner–Bratzler shear tests and texture profile analysis provide valuable information about the tenderness and sensory perception of meat, both tests suffer from a lack of standardization, which limits cross-study comparisons. Here, we provide guidelines to standardize meat testing and report meat stiffness as the single most informative mechanical parameter. Collecting big standardized data and sharing them with the community at large could empower researchers to harness the power of generative artificial intelligence to inform the systematic development of meat analogs with desired mechanical properties and functions, taste, and sensory perception.

Fungi-based meat is emerging as a promising class of nutritious, sustainable, and minimally processed biomaterials with the potential to complement or replace traditional animal and plant-based meats. However, its mechanical and sensory properties remain largely unknown. Here we characterize the quasi-static and dynamic mechanical behavior of fungi-based steak using multi-axial mechanical testing, rheology, and texture profile analysis. We find that the rate-independent response under quasi-static compression and shear is isotropic, while the rate-dependent response under dynamic compression is markedly anisotropic with stiffnesses and peak forces four times larger cross-plane than in-plane. Automated model discovery reveals that the exponential Demiray model best explains the rate-independent elastic response upon chewing in both directions. The rate-dependent directional stiffening can be linked, at least in part, to the high water content of mushroom root mycelium and to the restricted fluid flow at higher loading rates. Complementary sensory surveys reveal a strong correlation with mechanical metrics and suggest that we perceive the fungi-based steak as more moist, more viscous, and more fibrous than traditional animal- and plant-based meats. Taken together, our findings position fungi-based steak as an attractive, structurally equivalent, and sensorially compelling alternative protein source that is healthy for people and for the planet.