OpenNovelty.org

Model-Guided Microstimulation Steers Primate Visual Behavior

ICLR 2026 Conference SubmissionAnonymous Authors
OpenReview Score: 6.0Download Report PDF
causal interventionstopographic deep artificial neural networksbrain modeling
Abstract:

Brain stimulation is a powerful tool for understanding cortical function and holds the promise of therapeutic interventions to treat neuropsychiatric disorders such as impaired vision. Prototypical approaches to visual prosthetics apply patterns of electric microstimulation to the early visual cortex and can evoke percepts of simple symbols such as letters. However, these approaches are limited by the number of electrodes that can be implanted in early visual regions. Instead, higher-level visual regions are known to underlie the representations of complex visual objects such as faces and scenes and thus constitute a promising target for stimulating the cortex to elicit more complex visual experience. We developed a computational framework composed of two main components to address the challenge of stimulating cortex in high-dimensional object space spanned by higher-level visual cortex: 1. a causally predictive model that predicts primate behavior from image and stimulation input via topographic models and perturbation modules. 2. a mapping procedure that translates optimal model stimulation sites to monkey cortex. Testing our approach in two macaque monkeys that perform a visual recognition task, our results suggest that model-guided microstimulation is a promising approach to steer complex visual behavior. This proof-of-principle establishes a foundation for next-generation visual prosthetics that could restore complex visual experiences by stimulating higher-level visual cortex.

Disclaimer
This report is AI-GENERATED using Large Language Models and WisPaper (A scholar search engine). It analyzes academic papers' tasks and contributions against retrieved prior work. While this system identifies POTENTIAL overlaps and novel directions, ITS COVERAGE IS NOT EXHAUSTIVE AND JUDGMENTS ARE APPROXIMATE. These results are intended to assist human reviewers and SHOULD NOT be relied upon as a definitive verdict on novelty.
NOTE that some papers exist in multiple, slightly different versions (e.g., with different titles or URLs). The system may retrieve several versions of the same underlying work. The current automated pipeline does not reliably align or distinguish these cases, so human reviewers will need to disambiguate them manually.
If you have any questions, please contact: mingzhang23@m.fudan.edu.cn

Overview

Overall Novelty Assessment

The paper develops a computational framework to guide microstimulation of higher-level visual cortex, aiming to elicit complex visual percepts by targeting object-selective regions rather than early visual areas. Within the taxonomy, it occupies the 'Predictive Models for Behavior-Guided Stimulation' leaf under 'Computational Framework and Model-Guided Stimulation Approaches'. Notably, this leaf contains only the original paper itself—no sibling papers—indicating a sparse research direction within the broader field of six total papers across the taxonomy.

The taxonomy reveals three main branches: computational modeling approaches, empirical microstimulation studies in extrastriate cortex, and neural connectivity mechanisms. The paper's leaf sits alongside 'Neural Decoding for Visual Prosthetic Applications', which focuses on translating visual scenes into stimulation patterns for prosthetics. Neighboring branches include empirical work on optic flow perception, multisensory integration, and cortical mapping via electrical stimulation. The scope notes clarify that the paper's modeling-driven approach distinguishes it from purely exploratory empirical studies that lack computational prediction components.

Among twenty-one candidates examined, the first contribution—developing a computational framework for model-guided microstimulation—shows one refutable candidate out of three examined, suggesting some prior work in predictive modeling for stimulation. The second contribution, model-in-the-loop experimental validation in primates, examined ten candidates with none clearly refuting it, indicating relative novelty in closed-loop behavioral testing. The third contribution, image generation for visualizing stimulation effects, examined eight candidates with no refutations, suggesting this visualization approach may be less explored in prior literature.

Given the limited search scope of twenty-one semantically matched candidates, the analysis suggests the paper occupies a relatively sparse research direction, particularly in combining predictive modeling with primate behavioral validation. The absence of sibling papers in its taxonomy leaf and the low refutation rates for two of three contributions support this impression. However, the analysis does not cover exhaustive citation networks or domain-specific venues, leaving open the possibility of relevant work outside the top-K semantic matches examined.

Taxonomy

Core-task Taxonomy Papers
6
3
Claimed Contributions
21
Contribution Candidate Papers Compared
1
Refutable Paper

Research Landscape Overview

Core task: model-guided microstimulation of higher-level visual cortex. This field investigates how computational models can inform targeted electrical stimulation of extrastriate visual areas to probe and manipulate perceptual and cognitive processes. The taxonomy reveals three main branches. The first, Computational Framework and Model-Guided Stimulation Approaches, encompasses predictive models that link neural activity patterns to behavioral outcomes, enabling hypothesis-driven stimulation protocols. The second, Empirical Microstimulation Studies in Extrastriate Cortex, gathers direct experimental work applying microstimulation to areas beyond primary visual cortex—such as motion-sensitive regions studied in Optic Flow Roll[1] or multisensory integration sites explored in Cue Integration Physiology[2]—to causally test functional roles. The third branch, Neural Connectivity and Plasticity Mechanisms, examines how stimulation-induced changes propagate through cortical networks and reshape synaptic organization, often leveraging imaging techniques like Voltage Sensitive Dyes[5] or mapping methods seen in Eloquent Areas Mapping[3] and IMTES Visual Field[4]. Recent work highlights a tension between purely data-driven stimulation strategies and theory-driven approaches that use computational models to predict perceptual effects before intervention. Model-Guided Microstimulation[0] sits squarely within the predictive modeling cluster, emphasizing the use of quantitative frameworks to guide where and when to stimulate in order to elicit specific behavioral responses. This contrasts with more exploratory empirical studies that systematically probe extrastriate regions without strong a priori predictions, as well as with connectivity-focused investigations that prioritize understanding network-level consequences over immediate perceptual readouts. By integrating model predictions with causal manipulation, Model-Guided Microstimulation[0] bridges computational theory and experimental intervention, addressing the open question of how to optimally translate neural population codes into targeted stimulation parameters for higher-level visual areas.

Claimed Contributions

Computational framework for model-guided microstimulation of high-level visual cortex

The authors introduce a three-component computational framework that uses topographic deep neural networks with perturbation modules to simulate microstimulation effects, prototype experiments in silico, and map model predictions back to primate cortex for guiding causal interventions in higher-level visual areas.

3 retrieved papers
Can Refute
Model-in-the-loop experimental validation in primate visual behavior

The authors demonstrate that their framework can prospectively predict and induce behavioral shifts in macaque monkeys performing visual recognition tasks, with model predictions correlating with actual behavioral outcomes and producing significant in-vivo perceptual changes.

10 retrieved papers
Image generation method for visualizing perceptual consequences of stimulation

The authors develop visualization techniques using GAN-based and diffusion-based image generation to interpret the perceptual effects of simulated microstimulation, revealing face-like features emerging when stimulating face-selective model regions, analogous to reported facephenes in human patients.

8 retrieved papers

Core Task Comparisons

Comparisons with papers in the same taxonomy category

Within the taxonomy built over the current TopK core-task papers, the original paper is assigned to a leaf with no direct siblings and no cousin branches under the same grandparent topic. In this retrieved landscape, it appears structurally isolated, which is one partial signal of novelty, but still constrained by search coverage and taxonomy granularity.

Contribution Analysis

Detailed comparisons for each claimed contribution

Contribution

Computational framework for model-guided microstimulation of high-level visual cortex

The authors introduce a three-component computational framework that uses topographic deep neural networks with perturbation modules to simulate microstimulation effects, prototype experiments in silico, and map model predictions back to primate cortex for guiding causal interventions in higher-level visual areas.

Contribution

Model-in-the-loop experimental validation in primate visual behavior

The authors demonstrate that their framework can prospectively predict and induce behavioral shifts in macaque monkeys performing visual recognition tasks, with model predictions correlating with actual behavioral outcomes and producing significant in-vivo perceptual changes.

Contribution

Image generation method for visualizing perceptual consequences of stimulation

The authors develop visualization techniques using GAN-based and diffusion-based image generation to interpret the perceptual effects of simulated microstimulation, revealing face-like features emerging when stimulating face-selective model regions, analogous to reported facephenes in human patients.

Table of Contents