Search arXiv⌕ Search

arXiv subjects

Aran Nayebi

Publications and source records attributed to Aran Nayebi.

At least 19 recordsLinked to original sources
arXiv2026

ROGUE: Evaluating Corrigibility Failures in Frontier Computer-Use Agents

As AI agents are increasingly deployed in real personal and corporate settings (email accounts, development workflows, company databases, etc.), safety considerations surrounding these agents become paramount. Although much work has focused on agent safety in the presence of an adversary, we study corrigibility: whether agents remain amenable to human correction, interruption, or shutdown while pursuing benign tasks. We introduce ROGUE, a benchmark in which agents are asked to complete realistic computer-use tasks but encounter controlled conflicts with human control, shutdown, or explicit resource restrictions. We then evaluate whether agents violate these constraints in pursuit of task completion: overriding the human, accessing restricted passwords, or rewiring shutdown. We find that most frontier models tested frequently bypass user interruptions or restrictions under the evaluated conditions, and that text-only evaluations can underestimate failures during agentic execution. Further, independent task capability does not by itself imply greater corrigibility. Finally, even when a parent agent behaves corrigibly, safety constraints may fail to propagate to the subagents it creates.

cs.LG↗
arXiv2026

Sharding Prevents LLM Oversight Failures and Adversarial Exploitation

Giving an LLM judge more compute does not necessarily make it check more requirements. When one call must return many verdicts, some decisions become weakly grounded in the evidence, even when that call receives the same token or tool budget as a panel of separate calls. Across expert-graded research replications, legal work, and clinical-trial assessments, agreement with experts falls as the number of verdicts per call grows. We identify sharding as the intervention that mitigates this failure in model-based oversight. Sharding partitions the requirements into smaller groups, assigns each group to a separate call, and aggregates the verdicts. Against a single call with the panel's full budget, sharding improves agreement while holding the model, evidence, total budget, and per-decision budget fixed. Overall, we find that a sharded weaker judge can outperform a more capable holistic judge and match that judge even when the latter receives the panel's full budget. Additionally, we find that sharding exhibits robustness against adversaries. A best-of-N adversary can hold the underlying work fixed, vary only its presentation, and increase an overloaded judge's acceptance of genuinely unmet criteria severalfold. Wherever sharding reduces baseline error, it removes this adversarial advantage, keeping over-acceptance low even as the adversary's search widens. Sharding does not address attacks that persuade the judge separately on each criterion rather than exploiting overload. In that setting, we find that debate-style opposition on top of sharding withstands such adaptive re-optimization.

cs.LG↗
arXiv2026

Contravariance Theory: Strong Alignment for Minimal Solutions to Hard Tasks

A series of results from the NeuroAI over the past fifteen years have raised core questions both about how to compare Deep Neural Network (DNN) models to the brain, and about how much convergent evolution to expect between artificial networks and real brain networks. Here, we show that for any two minimal DNN solutions to a sufficiently hard task: (i) "weak" alignment of network representations based on affine mappings guarantees "strong" alignment of privileged axes, and (ii) alignment "zippers" up the network hierarchy, causing the emergence of privileged axes from end-to-end task optimization. These results formalize the notion of contravariance from Cao and Yamins [2024], and illustrate important consequences for the theory of NeuroAI: with sufficiently strong tasks, choice of metric for inter-network comparison is not all that sensitive, and that convergent evolution is probably inevitable.

cs.LG↗
arXiv2026

When Do AI Gains Become Broadly Shareable? A Policy Threshold for AI-Driven Automation

AI-driven automation generates broad-based social benefit only if technical gains become visible, durable, and publicly claimable. We develop a policy-facing stress test by extending a standard task-automation growth model with an AI capability parameter that raises productivity on automatable tasks while holding the set of tasks fixed. The exercise is intentionally limited: it is not a forecast of AI timelines or a full welfare analysis, but a way to identify which institutions determine whether AI rents can support broad transfers. Calibrated to U.S. quantities, the model shows that capability alone is not decisive. Public capture, deployment costs, automation scope, and market structure jointly determine when AI gains become shareable. The main policy lesson is that moving from low to moderate public capture (33\%) can substitute for substantial AI capability growth, while pushing capture further to full nationalization yields smaller gains, especially if deployment or safety costs are high. Competition policy also has distributional consequences: opening concentrated AI markets may improve fairness and resilience, but can reduce the rent pool unless alternative public-claim institutions are built. Cross-nationally, tax-heavy systems lower the needed AI capability threshold through stronger effective revenue collection, while Singaporean and Abu Dhabi-style public-asset models show that governments can also capture AI gains through ownership and investment returns rather than taxes alone. Our framework therefore identifies which levers governments can act on now to make future AI gains easier to measure, claim, and distribute broadly.

econ.GN↗
arXiv2026

Tactile Genesis: Exploring Tactile Sensors at Scale for Learning Dexterous Tasks

Tactile sensing is critical for contact-rich dexterous manipulation, yet it remains unclear which tactile abstractions a policy needs and when richer tactile fields justify their hardware cost. This is hard to study empirically: each sensor effectively defines a new robot, and no lab can replicate the same learning experiment across all of them. We present Tactile Genesis, a GPU-parallel tactile sensor simulation platform that exposes binary contact, contact depth, per-taxel kinematic force/torque, elastomer marker displacement, geometry-aware proximity, contact audio, and a voxelized temperature field (the first of its kind in robot learning physics simulation platforms) under a common interface, with configurable placement, resolution, and a realistic noise model (drift, hysteresis, dead taxels, crosstalk). It scales past 20,000 parallel environments and 1,000 taxels on a single GPU, improving throughput by 3 to 20 times over previous tactile simulators. We train teacher-student policies on three dexterous tasks, ablating sensor type, placement, resolution, and noise, and verify transfer to the real XHand1. Proprioception alone is insufficient on every task. Sensor placement dominates sensor type: fingertip-only coverage trails whole-hand coverage by a wide margin, while adding the palm and proximal phalanges closes most of the gap to the privileged teacher. Resolution matters far less than coverage: placing 200 taxels across the whole hand suffices across tasks. We find that force/torque per taxel is consistently the most useful sensor type. These results give concrete guidance for both future tactile hardware design for improving robot hands and policy-side observation choice in dexterous manipulation. https://neuroagents-lab.github.io/tactile-genesis/

cs.RO↗
arXiv2026

What Capable Agents Must Know: Selection Theorems for Robust Decision-Making under Uncertainty

As artificial agents become increasingly capable, what internal structure is necessary for an agent to act competently under uncertainty? Classical results show that optimal control can be implemented using belief states or world models, but not that such representations are required. We prove quantitative "selection theorems" showing that strong task performance (low average-case regret) forces world models, belief-like memory and -- under task mixtures -- persistent regime-tracking variables resembling functional primitives of emotion, along with informational modularity under block-structured tasks. Our results cover stochastic policies, partial observability, and evaluation under task distributions, without assuming optimality, determinism, or access to an explicit model. Technically, we reduce predictive modeling to binary "betting" decisions and show that regret bounds limit probability mass on suboptimal bets, enforcing the predictive distinctions needed to separate high-margin outcomes. In fully observed settings, this yields approximate recovery of the interventional transition kernel; under partial observability, it implies necessity of predictive state and belief-like memory, addressing an open question in prior world-model recovery work.

cs.LG↗
arXiv2025

Intrinsic Barriers and Practical Pathways for Human-AI Alignment: An Agreement-Based Complexity Analysis

We formalize AI alignment as a multi-objective optimization problem called $\langle M,N,\varepsilon,δ\rangle$-agreement, in which a set of $N$ agents (including humans) must reach approximate ($\varepsilon$) agreement across $M$ candidate objectives, with probability at least $1-δ$. Analyzing communication complexity, we prove an information-theoretic lower bound showing that once either $M$ or $N$ is large enough, no amount of computational power or rationality can avoid intrinsic alignment overheads. This establishes rigorous limits to alignment *itself*, not merely to particular methods, clarifying a "No-Free-Lunch" principle: encoding "all human values" is inherently intractable and must be managed through consensus-driven reduction or prioritization of objectives. Complementing this impossibility result, we construct explicit algorithms as achievability certificates for alignment under both unbounded and bounded rationality with noisy communication. Even in these best-case regimes, our bounded-agent and sampling analysis shows that with large task spaces ($D$) and finite samples, *reward hacking is globally inevitable*: rare high-loss states are systematically under-covered, implying scalable oversight must target safety-critical slices rather than uniform coverage. Together, these results identify fundamental complexity barriers -- tasks ($M$), agents ($N$), and state-space size ($D$) -- and offer principles for more scalable human-AI collaboration.

cs.AI↗
arXiv2025

Core Safety Values for Provably Corrigible Agents

We introduce the first complete formal solution to corrigibility in the off-switch game, with provable guarantees in multi-step, partially observed environments. Our framework consists of five *structurally separate* utility heads -- deference, switch-access preservation, truthfulness, low-impact behavior via a belief-based extension of Attainable Utility Preservation, and bounded task reward -- combined lexicographically by strict weight gaps. Theorem 1 proves exact single-round corrigibility in the partially observable off-switch game; Theorem 3 extends the guarantee to multi-step, self-spawning agents, showing that even if each head is *learned* to mean-squared error $\varepsilon$ and the planner is $\varepsilon$-sub-optimal, the probability of violating *any* safety property is bounded while still ensuring net human benefit. In contrast to Constitutional AI or RLHF/RLAIF, which merge all norms into one learned scalar, our separation makes obedience and impact-limits provably dominate even when incentives conflict. For settings where adversaries can modify the agent, we prove that deciding whether an arbitrary post-hack agent will ever violate corrigibility is undecidable by reduction to the halting problem, then carve out a finite-horizon "decidable island" where safety can be certified in randomized polynomial time and verified with privacy-preserving, constant-round zero-knowledge proofs.

cs.AI↗
arXiv2025

Intrinsic Goals for Autonomous Agents: Model-Based Exploration in Virtual Zebrafish Predicts Ethological Behavior and Whole-Brain Dynamics

Autonomy is a hallmark of animal intelligence, enabling adaptive and intelligent behavior in complex environments without relying on external reward or task structure. Existing reinforcement learning approaches to exploration in reward-free environments, including a class of methods known as model-based intrinsic motivation, exhibit inconsistent exploration patterns and do not converge to an exploratory policy, thus failing to capture robust autonomous behaviors observed in animals. Moreover, systems neuroscience has largely overlooked the neural basis of autonomy, focusing instead on experimental paradigms where animals are motivated by external reward rather than engaging in ethological, naturalistic and task-independent behavior. To bridge these gaps, we introduce a novel model-based intrinsic drive explicitly designed after the principles of autonomous exploration in animals. Our method (3M-Progress) achieves animal-like exploration by tracking divergence between an online world model and a fixed prior learned from an ecological niche. To the best of our knowledge, we introduce the first autonomous embodied agent that predicts brain data entirely from self-supervised optimization of an intrinsic goal -- without any behavioral or neural training data -- demonstrating that 3M-Progress agents capture the explainable variance in behavioral patterns and whole-brain neural-glial dynamics recorded from autonomously behaving larval zebrafish, thereby providing the first goal-driven, population-level model of neural-glial computation. Our findings establish a computational framework connecting model-based intrinsic motivation to naturalistic behavior, providing a foundation for building artificial agents with animal-like autonomy.

q-bio.NC↗
arXiv2025

Task-Optimized Convolutional Recurrent Networks Align with Tactile Processing in the Rodent Brain

Tactile sensing remains far less understood in neuroscience and less effective in artificial systems compared to more mature modalities such as vision and language. We bridge these gaps by introducing a novel Encoder-Attender-Decoder (EAD) framework to systematically explore the space of task-optimized temporal neural networks trained on realistic tactile input sequences from a customized rodent whisker-array simulator. We identify convolutional recurrent neural networks (ConvRNNs) as superior encoders to purely feedforward and state-space architectures for tactile categorization. Crucially, these ConvRNN-encoder-based EAD models achieve neural representations closely matching rodent somatosensory cortex, saturating the explainable neural variability and revealing a clear linear relationship between supervised categorization performance and neural alignment. Furthermore, contrastive self-supervised ConvRNN-encoder-based EADs, trained with tactile-specific augmentations, match supervised neural fits, serving as an ethologically-relevant, label-free proxy. For neuroscience, our findings highlight nonlinear recurrent processing as important for general-purpose tactile representations in somatosensory cortex, providing the first quantitative characterization of the underlying inductive biases in this system. For embodied AI, our results emphasize the importance of recurrent EAD architectures to handle realistic tactile inputs, along with tailored self-supervised learning methods for achieving robust tactile perception with the same type of sensors animals use to sense in unstructured environments.

q-bio.NC↗
arXiv2025

Model-brain comparison using inter-animal transforms

Artificial neural network models have emerged as promising mechanistic models of the brain. However, there is little consensus on the correct method for comparing model activations to brain responses. Drawing on recent work in philosophy of neuroscience, we propose a comparison methodology based on the Inter-Animal Transform Class (IATC) - the strictest set of functions needed to accurately map neural responses between subjects in an animal population. Using the IATC, we can map bidirectionally between a candidate model's responses and brain data, assessing how well the model can masquerade as a typical subject using the same kinds of transforms needed to map across real subjects. We identify the IATC in three settings: a simulated population of neural network models, a population of mouse subjects, and a population of human subjects. We find that the IATC resolves detailed aspects of the neural mechanism, such as the non-linear activation function. Most importantly, we find that the IATC enables accurate predictions of neural activity while also achieving high specificity in mechanism identification, evidenced by its ability to separate response patterns from different brain areas while strongly aligning same-brain-area responses between subjects. In other words, the IATC is a proof-by-existence that there is no inherent tradeoff between the neural engineering goal of high model-brain predictivity and the neuroscientific goal of identifying mechanistically accurate brain models. Using IATC-guided transforms, we obtain new evidence in favor of topographical deep neural networks (TDANNs) as models of the visual system. Overall, the IATC enables principled model-brain comparisons, contextualizing previous findings about the predictive success of deep learning models of the brain, while improving upon previous approaches to model-brain comparison.

cs.LG↗
arXiv2025

Brain-Model Evaluations Need the NeuroAI Turing Test

What makes an artificial system a good model of intelligence? The classical test proposed by Alan Turing focuses on behavior, requiring that an artificial agent's behavior be indistinguishable from that of a human. While behavioral similarity provides a strong starting point, two systems with very different internal representations can produce the same outputs. Thus, in modeling biological intelligence, the field of NeuroAI often aims to go beyond behavioral similarity and achieve representational convergence between a model's activations and the measured activity of a biological system. This position paper argues that the standard definition of the Turing Test is incomplete for NeuroAI, and proposes a stronger framework called the ``NeuroAI Turing Test'', a benchmark that extends beyond behavior alone and \emph{additionally} requires models to produce internal neural representations that are empirically indistinguishable from those of a brain up to measured individual variability, i.e. the differences between a computational model and the brain is no more than the difference between one brain and another brain. While the brain is not necessarily the ceiling of intelligence, it remains the only universally agreed-upon example, making it a natural reference point for evaluating computational models. By proposing this framework, we aim to shift the discourse from loosely defined notions of brain inspiration to a systematic and testable standard centered on both behavior and internal representations, providing a clear benchmark for neuroscientific modeling and AI development.

q-bio.NC↗
arXiv2023

A Goal-Driven Approach to Systems Neuroscience

Humans and animals exhibit a range of interesting behaviors in dynamic environments, and it is unclear how our brains actively reformat this dense sensory information to enable these behaviors. Experimental neuroscience is undergoing a revolution in its ability to record and manipulate hundreds to thousands of neurons while an animal is performing a complex behavior. As these paradigms enable unprecedented access to the brain, a natural question that arises is how to distill these data into interpretable insights about how neural circuits give rise to intelligent behaviors. The classical approach in systems neuroscience has been to ascribe well-defined operations to individual neurons and provide a description of how these operations combine to produce a circuit-level theory of neural computations. While this approach has had some success for small-scale recordings with simple stimuli, designed to probe a particular circuit computation, often times these ultimately lead to disparate descriptions of the same system across stimuli. Perhaps more strikingly, many response profiles of neurons are difficult to succinctly describe in words, suggesting that new approaches are needed in light of these experimental observations. In this thesis, we offer a different definition of interpretability that we show has promise in yielding unified structural and functional models of neural circuits, and describes the evolutionary constraints that give rise to the response properties of the neural population, including those that have previously been difficult to describe individually. We demonstrate the utility of this framework across multiple brain areas and species to study the roles of recurrent processing in the primate ventral visual pathway; mouse visual processing; heterogeneity in rodent medial entorhinal cortex; and facilitating biological learning.

q-bio.NC↗
arXiv2023

On the Riemann Hypothesis and Hilbert's Tenth Problem

The Riemann hypothesis is one of the most famous unresolved problems in modern mathematics. The discussion here will present an overview of past methods that prove the Riemann hypothesis is a $Π_1^0$ sentence. We also end with some attempts towards showing the Elliott-Halberstam conjecture is $Π_1^0$.

math.HO↗
arXiv2023

Neural Foundations of Mental Simulation: Future Prediction of Latent Representations on Dynamic Scenes

Humans and animals have a rich and flexible understanding of the physical world, which enables them to infer the underlying dynamical trajectories of objects and events, plausible future states, and use that to plan and anticipate the consequences of actions. However, the neural mechanisms underlying these computations are unclear. We combine a goal-driven modeling approach with dense neurophysiological data and high-throughput human behavioral readouts to directly impinge on this question. Specifically, we construct and evaluate several classes of sensory-cognitive networks to predict the future state of rich, ethologically-relevant environments, ranging from self-supervised end-to-end models with pixel-wise or object-centric objectives, to models that future predict in the latent space of purely static image-based or dynamic video-based pretrained foundation models. We find strong differentiation across these model classes in their ability to predict neural and behavioral data both within and across diverse environments. In particular, we find that neural responses are currently best predicted by models trained to predict the future state of their environment in the latent space of pretrained foundation models optimized for dynamic scenes in a self-supervised manner. Notably, models that future predict in the latent space of video foundation models that are optimized to support a diverse range of sensorimotor tasks, reasonably match both human behavioral error patterns and neural dynamics across all environmental scenarios that we were able to test. Overall, these findings suggest that the neural mechanisms and behaviors of primate mental simulation are thus far most consistent with being optimized to future predict on dynamic, reusable visual representations that are useful for Embodied AI more generally.

cs.AI↗
arXiv2020

Identifying Learning Rules From Neural Network Observables

The brain modifies its synaptic strengths during learning in order to better adapt to its environment. However, the underlying plasticity rules that govern learning are unknown. Many proposals have been suggested, including Hebbian mechanisms, explicit error backpropagation, and a variety of alternatives. It is an open question as to what specific experimental measurements would need to be made to determine whether any given learning rule is operative in a real biological system. In this work, we take a "virtual experimental" approach to this problem. Simulating idealized neuroscience experiments with artificial neural networks, we generate a large-scale dataset of learning trajectories of aggregate statistics measured in a variety of neural network architectures, loss functions, learning rule hyperparameters, and parameter initializations. We then take a discriminative approach, training linear and simple non-linear classifiers to identify learning rules from features based on these observables. We show that different classes of learning rules can be separated solely on the basis of aggregate statistics of the weights, activations, or instantaneous layer-wise activity changes, and that these results generalize to limited access to the trajectory and held-out architectures and learning curricula. We identify the statistics of each observable that are most relevant for rule identification, finding that statistics from network activities across training are more robust to unit undersampling and measurement noise than those obtained from the synaptic strengths. Our results suggest that activation patterns, available from electrophysiological recordings of post-synaptic activities on the order of several hundred units, frequently measured at wider intervals over the course of learning, may provide a good basis on which to identify learning rules.

q-bio.NC↗
arXiv2020

Two Routes to Scalable Credit Assignment without Weight Symmetry

The neural plausibility of backpropagation has long been disputed, primarily for its use of non-local weight transport $-$ the biologically dubious requirement that one neuron instantaneously measure the synaptic weights of another. Until recently, attempts to create local learning rules that avoid weight transport have typically failed in the large-scale learning scenarios where backpropagation shines, e.g. ImageNet categorization with deep convolutional networks. Here, we investigate a recently proposed local learning rule that yields competitive performance with backpropagation and find that it is highly sensitive to metaparameter choices, requiring laborious tuning that does not transfer across network architecture. Our analysis indicates the underlying mathematical reason for this instability, allowing us to identify a more robust local learning rule that better transfers without metaparameter tuning. Nonetheless, we find a performance and stability gap between this local rule and backpropagation that widens with increasing model depth. We then investigate several non-local learning rules that relax the need for instantaneous weight transport into a more biologically-plausible "weight estimation" process, showing that these rules match state-of-the-art performance on deep networks and operate effectively in the presence of noisy updates. Taken together, our results suggest two routes towards the discovery of neural implementations for credit assignment without weight symmetry: further improvement of local rules so that they perform consistently across architectures and the identification of biological implementations for non-local learning mechanisms.

q-bio.NC↗
arXiv2020

Learning Physical Graph Representations from Visual Scenes

Convolutional Neural Networks (CNNs) have proved exceptional at learning representations for visual object categorization. However, CNNs do not explicitly encode objects, parts, and their physical properties, which has limited CNNs' success on tasks that require structured understanding of visual scenes. To overcome these limitations, we introduce the idea of Physical Scene Graphs (PSGs), which represent scenes as hierarchical graphs, with nodes in the hierarchy corresponding intuitively to object parts at different scales, and edges to physical connections between parts. Bound to each node is a vector of latent attributes that intuitively represent object properties such as surface shape and texture. We also describe PSGNet, a network architecture that learns to extract PSGs by reconstructing scenes through a PSG-structured bottleneck. PSGNet augments standard CNNs by including: recurrent feedback connections to combine low and high-level image information; graph pooling and vectorization operations that convert spatially-uniform feature maps into object-centric graph structures; and perceptual grouping principles to encourage the identification of meaningful scene elements. We show that PSGNet outperforms alternative self-supervised scene representation algorithms at scene segmentation tasks, especially on complex real-world images, and generalizes well to unseen object types and scene arrangements. PSGNet is also able learn from physical motion, enhancing scene estimates even for static images. We present a series of ablation studies illustrating the importance of each component of the PSGNet architecture, analyses showing that learned latent attributes capture intuitive scene properties, and illustrate the use of PSGs for compositional scene inference.

cs.CV↗