We introduce AbBiBench (Antibody Binding Benchmarking), a benchmarking framework for antibody binding affinity maturation and design. Unlike existing antibody evaluation strategies that rely on antibody alone and its similarity to natural ones (e.g., amino acid identity rate, structural RMSD), AbBiBench considers an antibody-antigen (Ab-Ag) complex as a functional unit and evaluates the potential of an antibody design binding to given antigen by measuring protein model's likelihood on the Ab-Ag complex. We first curate, standardize, and share 9 datasets containing 9 antigens (involving influenza, anti-lysozyme, HER2, VEGF, integrin, and SARS-CoV-2) and 155,853 heavy chain mutated antibodies. Using these datasets, we systematically compare 14 protein models including masked language models, autoregressive language models, inverse folding models, diffusion-based generative models, and geometric graph models. The correlation between model likelihood and experimental affinity values is used to evaluate model performance. Additionally, in a case study to increase binding affinity of antibody F045-092 to antigen influenza H1N1, we evaluate the generative power of the top-performing models by sampling a set of new antibodies binding to the antigen and ranking them based on structural integrity and biophysical properties of the Ab-Ag complex. As a result, structure-conditioned inverse folding models outperform others in both affinity correlation and generation tasks. Overall, AbBiBench provides a unified, biologically grounded evaluation framework to facilitate the development of more effective, function-aware antibody design models.

While many neural networks focus on layers to process information, the GAIN model uses a grid-based structure to improve biological plausibility and the dynamics of the model. The grid structure helps neurons to interact with their closest neighbors and improve their connections with one another, which is seen in biological neurons. While also being implemented with the Izhikevich model this approach allows for a computationally efficient and biologically accurate simulation that can aid in the development of neural networks, large scale simulations, and the development in the neuroscience field. This adaptation of the Izhikevich model can improve the dynamics and accuracy of the model, allowing for its uses to be specialized but efficient.

Metapopulation models are commonly used in ecology, evolution, and epidemiology. These models usually entail homogeneity assumptions within patches and study networks of migration between patches to generate insights into conservation of species, differentiation of populations, and persistence of infectious diseases. Here, focusing on infectious disease epidemiology, we take a complementary approach and study the effects of individual variation within patches while neglecting any form of disease transmission between patches. Consistently with previous work on single populations, we show how metapopulation models that neglect in-patch heterogeneity also underestimate basic reproduction numbers ($\mathcal{R}_{0}$) and the effort required to control or eliminate infectious diseases by uniform interventions. We then go beyond this confirmatory result and introduce a scheme to infer distributions of individual susceptibility or exposure to infection based on suitable stratifications of a population into patches. We apply the resulting metapopulation models to a simple case study of the COVID-19 pandemic.

Gene set analysis (GSA) is a foundational approach for interpreting genomic data of diseases by linking genes to biological processes. However, conventional GSA methods overlook clinical context of the analyses, often generating long lists of enriched pathways with redundant, nonspecific, or irrelevant results. Interpreting these requires extensive, ad-hoc manual effort, reducing both reliability and reproducibility. To address this limitation, we introduce cGSA, a novel AI-driven framework that enhances GSA by incorporating context-aware pathway prioritization. cGSA integrates gene cluster detection, enrichment analysis, and large language models to identify pathways that are not only statistically significant but also biologically meaningful. Benchmarking on 102 manually curated gene sets across 19 diseases and ten disease-related biological mechanisms shows that cGSA outperforms baseline methods by over 30%, with expert validation confirming its increased precision and interpretability. Two independent case studies in melanoma and breast cancer further demonstrate its potential to uncover context-specific insights and support targeted hypothesis generation.

Mobile genetic elements (e.g., endogenous viruses, LINEs, SINEs) can transfer between genomes, even between species, triggering dramatic genetic changes. Endogenous viral elements (EVEs) arise when infectious viruses integrate into the host germline. EVEs integrate at specific sites; their genes or regulatory regions can be exapted and could induce chromosomal rearrangement. We propose that EVEs participate in adaptive radiations and that their parent viruses, through interspecific transfer, could initiate new species formation. By synchronously affecting multiple individuals, viral outbreaks generate shared genomic changes that both facilitate reproductive isolation and provide the simultaneous modifications needed for groups to emerge as founders of new species. We suggest horizontal viral transfer during the K-Pg accelerated mammalian radiation linking viral epidemics to macroevolutionary diversification. This theoretical work proposes endogenous viruses as catalysts for explosive speciation.

Medical data range from genomic sequences and retinal photographs to structured laboratory results and unstructured clinical narratives. Although these modalities appear disparate, many encode convergent information about a single underlying physiological state. The Latent Space Hypothesis frames each observation as a projection of a unified, hierarchically organized manifold -- much like shadows cast by the same three-dimensional object. Within this learned geometric representation, an individual's health status occupies a point, disease progression traces a trajectory, and therapeutic intervention corresponds to a directed vector. Interpreting heterogeneous evidence in a shared space provides a principled way to re-examine eponymous conditions -- such as Parkinson's or Crohn's -- that often mask multiple pathophysiological entities and involve broader anatomical domains than once believed. By revealing sub-trajectories and patient-specific directions of change, the framework supplies a quantitative rationale for personalised diagnosis, longitudinal monitoring, and tailored treatment, moving clinical practice away from grouping by potentially misleading labels toward navigation of each person's unique trajectory. Challenges remain -- bias amplification, data scarcity for rare disorders, privacy, and the correlation-causation divide -- but scale-aware encoders, continual learning on longitudinal data streams, and perturbation-based validation offer plausible paths forward.

Computational inverse problems for biomedical simulators suffer from limited data and relatively high parameter dimensionality. This often requires sensitivity analysis, where parameters of the model are ranked based on their influence on the specific quantities of interest. This is especially important for simulators used to build medical digital twins, as the amount of data is typically limited. For expensive models, such as blood flow models, emulation is employed to expedite the simulation time. Parameter ranking and fixing using sensitivity analysis are often heuristic, though, and vary with the specific application or simulator used. The present study provides an innovative solution to this problem by leveraging polynomial chaos expansions (PCEs) for both multioutput global sensitivity analysis and formal parameter identifiability. For the former, we use dimension reduction to efficiently quantify time-series sensitivity of a one-dimensional pulmonary hemodynamics model. We consider both Windkessel and structured tree boundary conditions. We then use PCEs to construct profile-likelihood confidence intervals to formally assess parameter identifiability, and show how changes in experimental design improve identifiability. Our work presents a novel approach to determining parameter identifiability and leverages a common emulation strategy for enabling profile-likelihood analysis in problems governed by partial differential equations.

Despite a strong desire to quit, individuals with long-term substance use disorder (SUD) often struggle to resist drug use, even when aware of its harmful consequences. This disconnect between knowledge and compulsive behavior reflects a fundamental cognitive-behavioral conflict in addiction. Neurobiologically, differential cue-induced activity within striatal subregions, along with dopamine-mediated connectivity from the ventral to the dorsal striatum, contributes to compulsive drug-seeking. However, the functional mechanism linking these findings to behavioral conflict remains unclear. Another hallmark of addiction is temporal discounting: individuals with drug dependence exhibit steeper discount rates than non-users. Assuming the ventral-dorsal striatal organization reflects a gradient from cognitive to motor representations, addiction can be modeled within a hierarchical reinforcement learning (HRL) framework. However, integrating discounting into biologically grounded HRL remains an open challenge. In this work, we build on a model showing how action choices reinforced with drug rewards become insensitive to the negative consequences that follow. We address the integration of discounting by ensuring natural reward values converge across all levels in the HRL hierarchy, while drug rewards diverge due to their dopaminergic effects. Our results show that high discounting amplifies drug-seeking across the hierarchy, linking faster discounting with increased addiction severity and impulsivity. We demonstrate alignment with empirical findings on temporal discounting and propose testable predictions, establishing addiction as a disorder of hierarchical decision-making.

We study single cell E.coli chemotaxis in a spatio-temporally varying attractant environment. Modeling the attractant concentration in the form of a traveling sine wave, we measure in our simulations, the chemotactic drift velocity of the cell for different propagation speed of the attractant wave. We find a highly non-trivial dependence where the chemotactic drift velocity changes sign, and also shows multiple peaks. For slowly moving attractant wave, drift velocity is negative, i.e. the drift motion is directed opposite to wave propagation. As the wave speed increases, drift velocity shows a negative peak, then changes sign, reaches a positive peak and finally becomes zero when the wave moves too fast for the cell to respond. We explain this rich behavior from the difference in attractant gradient perceived by the cell during its run along the propagation direction and opposite to it. In particular, when the cell moves in the same direction as the wave, the relative velocity of the cell with respect to the wave becomes zero when the wave speed matches the run speed. In this limit, the cell is able to ride the wave and experiences no concentration gradient during these runs. On the contrary, for runs in the opposite direction, no such effect is present and the effective gradient increases monotonically with the wave speed. We show, using detailed quantitative measurements, how this difference gives rise to the counter-intuitive behavior of chemotactic drift velocity described above.

The measurement-induced phase transition (MIPT) is a recently formulated phenomenon in out-of-equilibrium systems. The competition between unitary evolutions and measurement-induced non-unitaries leads to the transition between the entangled and disentangled phases at some critical measurement rate. We conjecture that self-organized MIPT plays a key role in the generative model of cognitive networks and the formation of the state of consciousness in the "newborn-adult" transition. To this aim, we formulate the probe-target picture for the brain and suggest that MIPT interpreted as learnability transition takes place in the mental part of the target where the sites in the cognitive networks of semantic memory are concepts. Comparison with the synchronization phase transitions in the probe is made.

Modern neural recording techniques such as two-photon imaging allow to acquire vast time-series datasets with responses of hundreds or thousands of neurons. Contrastive learning is a powerful self-supervised framework for learning representations of complex datasets. Existing applications for neural time series rely on generic data augmentations and do not exploit the multi-trial data structure inherent in many neural datasets. Here we present TRACE, a new contrastive learning framework that averages across different subsets of trials to generate positive pairs. TRACE allows to directly learn a two-dimensional embedding, combining ideas from contrastive learning and neighbor embeddings. We show that TRACE outperforms other methods, resolving fine response differences in simulated data. Further, using in vivo recordings, we show that the representations learned by TRACE capture both biologically relevant continuous variation, cell-type-related cluster structure, and can assist data quality control.

The hippocampus encodes space through a striking gradient of place field sizes along its dorsal-ventral axis, yet the principles generating this continuous gradient from discrete grid cell inputs remain debated. We propose a unified theoretical framework establishing that hippocampal place fields arise naturally as linear projections of grid cell population activity, interpretable as eigenmodes. Critically, we demonstrate that a frequency-dependent decay of these grid-to-place connection weights naturally transforms inputs from discrete grid modules into a continuous spectrum of place field sizes. This multiscale organization is functionally significant: we reveal it shapes the inductive bias of the population code, balancing a fundamental trade-off between precision and generalization. Mathematical analysis and simulations demonstrate an optimal place field size for few-shot learning, which scales with environment structure. Our results offer a principled explanation for the place field gradient and generate testable predictions, bridging anatomical connectivity with adaptive learning in both biological and artificial intelligence.

Real-time decoding of neural activity is central to neuroscience and neurotechnology applications, from closed-loop experiments to brain-computer interfaces, where models are subject to strict latency constraints. Traditional methods, including simple recurrent neural networks, are fast and lightweight but often struggle to generalize to unseen data. In contrast, recent Transformer-based approaches leverage large-scale pretraining for strong generalization performance, but typically have much larger computational requirements and are not always suitable for low-resource or real-time settings. To address these shortcomings, we present POSSM, a novel hybrid architecture that combines individual spike tokenization via a cross-attention module with a recurrent state-space model (SSM) backbone to enable (1) fast and causal online prediction on neural activity and (2) efficient generalization to new sessions, individuals, and tasks through multi-dataset pretraining. We evaluate POSSM's decoding performance and inference speed on intracortical decoding of monkey motor tasks, and show that it extends to clinical applications, namely handwriting and speech decoding in human subjects. Notably, we demonstrate that pretraining on monkey motor-cortical recordings improves decoding performance on the human handwriting task, highlighting the exciting potential for cross-species transfer. In all of these tasks, we find that POSSM achieves decoding accuracy comparable to state-of-the-art Transformers, at a fraction of the inference cost (up to 9x faster on GPU). These results suggest that hybrid SSMs are a promising approach to bridging the gap between accuracy, inference speed, and generalization when training neural decoders for real-time, closed-loop applications.

Cooperation is vital for the survival of living systems but is challenging due to the costs borne by altruistic individuals. Direct reciprocity, where actions are based on past encounters, is a key mechanism fostering cooperation. However, most studies assume synchronous decision-making, whereas real-world interactions are often asynchronous, with individuals acting in sequence. This asynchrony can undermine standard cooperative strategies like Tit-for-Tat and Win-Stay Lose-Shift. To better understand cooperation in real-world contexts, it is crucial to explore the theory of direct reciprocity in asynchronous interactions. To address this, we introduce a framework based on asynchronous stochastic games, incorporating asynchronous decisions and dynamic environmental feedback. We analytically derive the conditions under which strategies form cooperative Nash equilibria. Our results demonstrate that the order of interactions can significantly alter outcomes: interaction asynchrony generally inhibits cooperation, except under specific conditions where environmental feedback effectively mitigates its negative impact. When environmental feedback is incorporated, a variety of stable reciprocal strategies can be sustained. Notably, above a critical environmental threshold, any cooperative strategy can form a Nash equilibrium. Overall, our work underscores the importance of interaction order in long-term evolutionary processes and highlights the pivotal role of environmental feedback in stabilizing cooperation in asynchronous interactions.

Transformer-based models have demonstrated remarkable reasoning abilities, but the mechanisms underlying relational reasoning in different learning regimes remain poorly understood. In this work, we investigate how transformers perform a classic relational reasoning task from the Psychology literature, \textit{transitive inference}, which requires inference about indirectly related items by integrating information across observed adjacent item pairs (e.g., if A>B and B>C, then A>C). We compare transitive inference behavior across two distinct learning regimes: in-weights learning (IWL), where models store information in network parameters, and in-context learning (ICL), where models flexibly utilize information presented within the input sequence. Our findings reveal that IWL naturally induces a generalization bias towards transitive inference, despite being trained only on adjacent items, whereas ICL models trained solely on adjacent items do not generalize transitively. Mechanistic analysis shows that ICL models develop induction circuits that implement a simple match-and-copy strategy that performs well at relating adjacent pairs, but does not encoding hierarchical relationships among indirectly related items. Interestingly, when pre-trained on in-context linear regression tasks, transformers successfully exhibit in-context generalizable transitive inference. Moreover, like IWL, they display both \textit{symbolic distance} and \textit{terminal item effects} characteristic of human and animal performance, without forming induction circuits. These results suggest that pre-training on tasks with underlying structure promotes the development of representations that can scaffold in-context relational reasoning.

Deep neural networks (DNNs) trained on visual tasks develop feature representations that resemble those in the human visual system. Although DNN-based encoding models can accurately predict brain responses to visual stimuli, they offer limited insight into the specific features driving these responses. Here, we demonstrate that activation maximization -- a technique designed to interpret vision DNNs -- can be applied to DNN-based encoding models of the human brain. We extract and adaptively downsample activations from multiple layers of a pretrained Inception V3 network, then use linear regression to predict fMRI responses. This yields a full image-computable model of brain responses. Next, we apply activation maximization to generate images optimized for predicted responses in individual cortical voxels. We find that these images contain visual characteristics that qualitatively correspond with known selectivity and enable exploration of selectivity across the visual cortex. We further extend our method to whole regions of interest (ROIs) of the brain and validate its efficacy by presenting these images to human participants in an fMRI study. We find that the generated images reliably drive activity in targeted regions across both low- and high-level visual areas and across subjects. These results demonstrate that activation maximization can be successfully applied to DNN-based encoding models. By addressing key limitations of alternative approaches that require natively generative models, our approach enables flexible characterization and modulation of responses across the human visual system.

Protein structure prediction models are now capable of generating accurate 3D structural hypotheses from sequence alone. However, they routinely fail to capture the conformational diversity of dynamic biomolecular complexes, often requiring heuristic MSA subsampling approaches for generating alternative states. In parallel, cryo-electron microscopy (cryo-EM) has emerged as a powerful tool for imaging near-native structural heterogeneity, but is challenged by arduous pipelines to go from raw experimental data to atomic models. Here, we bridge the gap between these modalities, combining cryo-EM density maps with the rich sequence and biophysical priors learned by protein structure prediction models. Our method, CryoBoltz, guides the sampling trajectory of a pretrained protein structure prediction model using both global and local structural constraints derived from density maps, driving predictions towards conformational states consistent with the experimental data. We demonstrate that this flexible yet powerful inference-time approach allows us to build atomic models into heterogeneous cryo-EM maps across a variety of dynamic biomolecular systems including transporters and antibodies.

Characterizing the diverse computational properties of human neurons via multimodal electrophysiological, transcriptomic, and morphological data provides the foundation for constructing and validating bio-realistic neuron models that can advance our understanding of fundamental mechanisms underlying brain function. However, current modeling approaches remain constrained by the limited availability and intrinsic variability of experimental neuronal data. To capture variability, ensembles of deterministic models are often used, but are difficult to scale as model generation requires repeating computationally expensive optimization for each neuron. While deep learning is becoming increasingly relevant in this space, it fails to capture the full biophysical complexity of neurons, their nonlinear voltage dynamics, and variability. To address these shortcomings, we introduce NOBLE, a neural operator framework that learns a mapping from a continuous frequency-modulated embedding of interpretable neuron features to the somatic voltage response induced by current injection. Trained on data generated from biophysically realistic neuron models, NOBLE predicts distributions of neural dynamics accounting for the intrinsic experimental variability. Unlike conventional bio-realistic neuron models, interpolating within the embedding space offers models whose dynamics are consistent with experimentally observed responses. NOBLE is the first scaled-up deep learning framework validated on real experimental data, enabling efficient generation of synthetic neurons that exhibit trial-to-trial variability and achieve a $4200\times$ speedup over numerical solvers. To this end, NOBLE captures fundamental neural properties, opening the door to a better understanding of cellular composition and computations, neuromorphic architectures, large-scale brain circuits, and general neuroAI applications.

The digitization of histology slides has revolutionized pathology, providing massive datasets for cancer diagnosis and research. Contrastive self-supervised and vision-language models have been shown to effectively mine large pathology datasets to learn discriminative representations. On the other hand, generative models, capable of synthesizing realistic and diverse images, present a compelling solution to address unique problems in pathology that involve synthesizing images; overcoming annotated data scarcity, enabling privacy-preserving data sharing, and performing inherently generative tasks, such as virtual staining. We introduce PixCell, the first diffusion-based generative foundation model for histopathology. We train PixCell on PanCan-30M, a vast, diverse dataset derived from 69,184 H\&E-stained whole slide images covering various cancer types. We employ a progressive training strategy and a self-supervision-based conditioning that allows us to scale up training without any annotated data. PixCell generates diverse and high-quality images across multiple cancer types, which we find can be used in place of real data to train a self-supervised discriminative model. Synthetic images shared between institutions are subject to fewer regulatory barriers than would be the case with real clinical images. Furthermore, we showcase the ability to precisely control image generation using a small set of annotated images, which can be used for both data augmentation and educational purposes. Testing on a cell segmentation task, a mask-guided PixCell enables targeted data augmentation, improving downstream performance. Finally, we demonstrate PixCell's ability to use H\&E structural staining to infer results from molecular marker studies; we use this capability to infer IHC staining from H\&E images. Our trained models are publicly released to accelerate research in computational pathology.

Connections between generative diffusion and continuous-state associative memory models are studied. Morse-Smale dynamical systems are emphasized as universal approximators of gradient-based associative memory models and diffusion models as white-noise perturbed systems thereof. Universal properties of associative memory that follow from this description are described and used to characterize a generic transition from generation to memory as noise levels diminish. Structural stability inherited by Morse-Smale flows is shown to imply a notion of stability for diffusions at vanishing noise levels. Applied to one- and two-parameter families of gradients, this indicates stability at all but isolated points of associative memory learning landscapes and the learning and generation landscapes of diffusion models with gradient drift in the zero-noise limit, at which small sets of generic bifurcations characterize qualitative transitions between stable systems. Examples illustrating the characterization of these landscapes by sequences of these bifurcations are given, along with structural stability criterion for classic and modern Hopfield networks (equivalently, the attention mechanism).

Associative memory models such as the Hopfield network and its dense generalizations with higher-order interactions exhibit a "blackout catastrophe"--a discontinuous transition where stable memory states abruptly vanish when the number of stored patterns exceeds a critical capacity. This transition is often interpreted as rendering networks unusable beyond capacity limits. We argue that this interpretation is largely an artifact of the equilibrium perspective. We derive dynamical mean-field equations using a bipartite cavity approach for graded-activity dense associative memory models, with the Hopfield model as a special case, and solve them using a numerical scheme. We show that patterns can be transiently retrieved with high accuracy above capacity despite the absence of stable attractors. This occurs because slow regions persist in the above-capacity energy landscape as shallow, unstable remnants of below-capacity stable basins. The same transient-retrieval effect occurs in below-capacity networks initialized outside basins of attraction. "Transient-recovery curves" provide a concise visual summary of these effects, revealing graceful, non-catastrophic changes in retrieval behavior above capacity and allowing us to compare the behavior across interaction orders. This dynamical perspective reveals rich energy landscape structure obscured by equilibrium analysis and suggests biological neural circuits may exploit transient dynamics for memory retrieval. Furthermore, our approach suggests ways of understanding computational properties of neural circuits without reference to fixed points, advances the technical repertoire of numerical mean-field solution methods for recurrent neural networks, and yields new theoretical results on generalizations of the Hopfield model.