Jeffrey L. Krichmar
Current Organizations / Institutes
The organizations/institutes from the scholar's latest article
Previous Organizations / Institutes
The organizations / institutes from the scholar's other articles in the past ten years
Research Subjects
Jeffrey L. Krichmar has a total of 284 co-authors. Here are the top co-authors by number of shared publications:
- Nikil Dutt
University of California at Irvine
19 shared publications - Derrik E. Asher
DEVCOM Army Research Laboratory, USA
7 shared publications Gerald M. Edelman
The Neurosciences Institute
7 shared publications- Praveen Kumar Pilly
Intelligent Systems Center, HRL Laboratories, 3011 Malibu Canyon Road, Malibu, 90265, CA, USA
7 shared publications - Xinyun Zou
Division of Biology and Biological Engineering California Institute of Technology 91125 Pasadena CA USA
7 shared publications Andrew Zaldivar
Department of Cognitive Sciences, University of California, Irvine Irvine, CA, USA
6 shared publications
Latest Publications
- 1 October 2026
- Published by Springer Nature
- p. 335-346
Tactile Discrimination in a Soft Octopus-Inspired Robot
- 1 October 2026
- Published by Springer Nature
- p. 335-346
Tactile Discrimination in a Soft Octopus-Inspired Robot
- 1 October 2026
- Published by Springer Nature
- p. 153-164
Delayed Gratification in a Robotic Model of Cuttlefish Behavior
- 1 October 2026
- Published by Springer Nature
- p. 153-164
Delayed Gratification in a Robotic Model of Cuttlefish Behavior
- 1 October 2026
- Published by Springer Nature
- p. 115-126
SYNCOFLY: An Interactive Platform for Collective Rhythm Synchronization
- 1 October 2026
- Published by Springer Nature
- p. 115-126
SYNCOFLY: An Interactive Platform for Collective Rhythm Synchronization
- Open Access
- 13 December 2025
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Episodic-like memory in a simulation of cuttlefish behavior
Episodic memory involves remembering the what, when, and where components of an event. It has been observed in humans, other vertebrates, and the invertebrate cuttlefish. In clever behavioral experiments, cuttlefish have been shown to have episodic-like memory, where they demonstrate the ability to remember when and where a preferred food source will appear. The present work replicates this behavior with a parsimonious model of episodic memory. To further test this model and explore episodic-like memory, we introduce a predator-prey scenario in which the agent must remember what creatures (e.g. predator, desirable prey, or less desirable prey) appear at a given time and region of the model environment. This simulates similar situations that cuttlefish face in the wild. They will typically hide when predators are in the area, and hunt for prey when available. When the memory model is queried for an action (e.g., hunt or hide), the cuttlefish agent hunts for preferred food, like shrimp, when available, and hides at other times when a predator appears. When the memory model is queried for a place, the cuttlefish agent acts opportunistically, seeking less-preferred food (e.g., crabs) if it is located farther from a predator. These differences show how behavior can be altered depending on how memory is accessed. Querying the model over time might mimic mental time travel, a hallmark of episodic memory. Although developed with cuttlefish in mind, the model shares similarities with the hippocampal indexing theory and captures aspects of vertebrate episodic memory. This suggests that the underlying mechanisms supporting episodic-like behavior in the present model may be an example of convergent cognitive evolution.
- 13 December 2025
- Published by Springer Nature in Scientific Reports
- Vol. 16 (1)
Open AccessEpisodic-like memory in a simulation of cuttlefish behavior
Episodic memory involves remembering the what, when, and where components of an event. It has been observed in humans, other vertebrates, and the invertebrate cuttlefish. In clever behavioral experiments, cuttlefish have been shown to have episodic-like memory, where they demonstrate the ability to remember when and where a preferred food source will appear. The present work replicates this behavior with a parsimonious model of episodic memory. To further test this model and explore episodic-like memory, we introduce a predator-prey scenario in which the agent must remember what creatures (e.g. predator, desirable prey, or less desirable prey) appear at a given time and region of the model environment. This simulates similar situations that cuttlefish face in the wild. They will typically hide when predators are in the area, and hunt for prey when available. When the memory model is queried for an action (e.g., hunt or hide), the cuttlefish agent hunts for preferred food, like shrimp, when available, and hides at other times when a predator appears. When the memory model is queried for a place, the cuttlefish agent acts opportunistically, seeking less-preferred food (e.g., crabs) if it is located farther from a predator. These differences show how behavior can be altered depending on how memory is accessed. Querying the model over time might mimic mental time travel, a hallmark of episodic memory. Although developed with cuttlefish in mind, the model shares similarities with the hippocampal indexing theory and captures aspects of vertebrate episodic memory. This suggests that the underlying mechanisms supporting episodic-like behavior in the present model may be an example of convergent cognitive evolution.
- 11 September 2025
- Published by SAGE Publications in Adaptive Behavior
- Vol. 33 (5-6) , 273-277
From Animals to Animats: The Tradition Continues
The Simulation of Adaptive Behavior (SAB) conference has a long tradition of bringing together researchers in computer science, artificial intelligence, artificial life, control, robotics, neurosciences, ethology, and evolutionary biology. Starting in 1990 and meeting every 2 years, it recently convened the 17 th meeting in Irvine, California, USA. The goal of this meeting, which is called From Animals to Animats , is to further our understanding of the behaviors and underlying mechanisms that allow natural and artificial animals to adapt and survive in uncertain environments. We asked several contributors of highly regarded presentations to submit an extended version of their work to this Special Issue of Adaptive Behavior. The articles included in this volume reflect the interdisciplinary nature of SAB, as well as the overarching theme of SAB to draw inspiration from biology.
- 11 September 2025
- Published by SAGE Publications in Adaptive Behavior
- Vol. 33 (5-6) , 273-277
From Animals to Animats: The Tradition Continues
The Simulation of Adaptive Behavior (SAB) conference has a long tradition of bringing together researchers in computer science, artificial intelligence, artificial life, control, robotics, neurosciences, ethology, and evolutionary biology. Starting in 1990 and meeting every 2 years, it recently convened the 17 th meeting in Irvine, California, USA. The goal of this meeting, which is called From Animals to Animats , is to further our understanding of the behaviors and underlying mechanisms that allow natural and artificial animals to adapt and survive in uncertain environments. We asked several contributors of highly regarded presentations to submit an extended version of their work to this Special Issue of Adaptive Behavior. The articles included in this volume reflect the interdisciplinary nature of SAB, as well as the overarching theme of SAB to draw inspiration from biology.
- Open Access
Episodic-Like Memory in a Simulation of Cuttlefish Behavior *
Episodic memory involves remembering the what, when, and where components of an event. It has been observed in humans, other vertebrates, and cephalopods. In clever behavioral experiments, cuttlefish have been shown to have episodic-like memory, where they demonstrate the ability to remember when and where a preferred food source will appear. The present work replicates this behavior with a parsimonious model of episodic memory. To further test this model and explore episodic-like memory, we introduce a predator-prey scenario in which the agent must remember what creatures (e.g. predator, desirable prey, or less desirable prey) appear at a given time and region of the model environment. This simulates similar situations that cephalopods face in the wild. They will typically hide when predators are in the area, and hunt for prey when available. When the memory model is queried for an action (e.g., hunt or hide), the cuttlefish agent hunts for preferred food, like shrimp, when available, and hides at other times when a predator appears. When the memory model is queried for a place, the cuttlefish agent acts opportunistically, seeking less-preferred food (e.g., crabs) if it is located farther from a predator. These differences show how behavior can be altered depending on how memory is accessed. Querying the model over time might mimic mental time travel, a hallmark of episodic memory. Although developed with cephalopods in mind, the model shares similarities with the hippocampal indexing theory and captures aspects of vertebrate episodic memory. This suggests that the underlying mechanisms supporting episodic-like behavior in the present model may not be unique to cephalopods.
Open AccessEpisodic-Like Memory in a Simulation of Cuttlefish Behavior *
Episodic memory involves remembering the what, when, and where components of an event. It has been observed in humans, other vertebrates, and cephalopods. In clever behavioral experiments, cuttlefish have been shown to have episodic-like memory, where they demonstrate the ability to remember when and where a preferred food source will appear. The present work replicates this behavior with a parsimonious model of episodic memory. To further test this model and explore episodic-like memory, we introduce a predator-prey scenario in which the agent must remember what creatures (e.g. predator, desirable prey, or less desirable prey) appear at a given time and region of the model environment. This simulates similar situations that cephalopods face in the wild. They will typically hide when predators are in the area, and hunt for prey when available. When the memory model is queried for an action (e.g., hunt or hide), the cuttlefish agent hunts for preferred food, like shrimp, when available, and hides at other times when a predator appears. When the memory model is queried for a place, the cuttlefish agent acts opportunistically, seeking less-preferred food (e.g., crabs) if it is located farther from a predator. These differences show how behavior can be altered depending on how memory is accessed. Querying the model over time might mimic mental time travel, a hallmark of episodic memory. Although developed with cephalopods in mind, the model shares similarities with the hippocampal indexing theory and captures aspects of vertebrate episodic memory. This suggests that the underlying mechanisms supporting episodic-like behavior in the present model may not be unique to cephalopods.
- Open Access
- 21 August 2025
- Published by SAGE Publications in Adaptive Behavior
- Vol. 33 (5-6) , 279-290
Adaptive, Goal-Oriented Navigation Using a Model of Directionally-Polarized Place Cells
CA1 place cells in the hippocampus have been shown to exhibit directional tuning properties, forming vector fields pointing towards locations in the environment known as ConSinks (Ormond & O’Keefe, 2022). We present a model, inspired by these findings, for learning goal-oriented navigation tasks. Our model employs a population of place cells that develop directional preferences, and are updated via a novel reward-modulated learning rule that refines directional turning of individual cells based on experience. Agents using this model navigated to goals significantly faster and more reliably than state-of-the-art Reinforcement Learning algorithms such as Deep Q-Networks (DQN) and Proximal Policy Optimization (PPO). We also demonstrate adaptation to new goals in a manner consistent with experimental findings, where the mean ConSink location shifts towards the new goal after it is introduced. Further experiments show that the model performs well with both goal-directed and random initialization of directional sensitivity, and that place cell density enhances learning efficiency. These results suggest a functional role for directional place cells in complex and obstacle filled environments.
- 21 August 2025
- Published by SAGE Publications in Adaptive Behavior
- Vol. 33 (5-6) , 279-290
Open AccessAdaptive, Goal-Oriented Navigation Using a Model of Directionally-Polarized Place Cells
CA1 place cells in the hippocampus have been shown to exhibit directional tuning properties, forming vector fields pointing towards locations in the environment known as ConSinks (Ormond & O’Keefe, 2022). We present a model, inspired by these findings, for learning goal-oriented navigation tasks. Our model employs a population of place cells that develop directional preferences, and are updated via a novel reward-modulated learning rule that refines directional turning of individual cells based on experience. Agents using this model navigated to goals significantly faster and more reliably than state-of-the-art Reinforcement Learning algorithms such as Deep Q-Networks (DQN) and Proximal Policy Optimization (PPO). We also demonstrate adaptation to new goals in a manner consistent with experimental findings, where the mean ConSink location shifts towards the new goal after it is introduced. Further experiments show that the model performs well with both goal-directed and random initialization of directional sensitivity, and that place cell density enhances learning efficiency. These results suggest a functional role for directional place cells in complex and obstacle filled environments.
- Open Access
STRMs: Spatial Temporal Reasoning Models for Vision-Based Localization Rivaling GPS Precision
This paper explores vision-based localization through a biologically-inspired approach that mirrors how humans and animals link views or perspectives when navigating their world. We introduce two sequential generative models, VAE-RNN and VAE-Transformer, which transform first-person perspective (FPP) observations into global map perspective (GMP) representations and precise geographical coordinates. Unlike retrieval-based methods, our approach frames localization as a generative task, learning direct mappings between perspectives without relying on dense satellite image databases. We evaluate these models across two real-world environments: a university campus navigated by a Jackal robot and an urban downtown area navigated by a Tesla sedan. The VAE-Transformer achieves impressive precision, with median deviations of 2.29m (1.37% of environment size) and 4.45m (0.35% of environment size) respectively, outperforming both VAE-RNN and prior cross-view geo-localization approaches. Our comprehensive Localization Performance Characteristics (LPC) analysis demonstrates superior performance with the VAE-Transformer achieving an AUC of 0.777 compared to 0.295 for VIGOR 200 and 0.225 for TransGeo, establishing a new state-of-the-art in vision-based localization. In some scenarios, our vision-based system rivals commercial smartphone GPS accuracy (AUC of 0.797) while requiring 5x less GPU memory and delivering 3x faster inference than existing methods in cross-view geo-localization. These results demonstrate that models inspired by biological spatial navigation can effectively memorize complex, dynamic environments and provide precise localization with minimal computational resources.
Open AccessSTRMs: Spatial Temporal Reasoning Models for Vision-Based Localization Rivaling GPS Precision
This paper explores vision-based localization through a biologically-inspired approach that mirrors how humans and animals link views or perspectives when navigating their world. We introduce two sequential generative models, VAE-RNN and VAE-Transformer, which transform first-person perspective (FPP) observations into global map perspective (GMP) representations and precise geographical coordinates. Unlike retrieval-based methods, our approach frames localization as a generative task, learning direct mappings between perspectives without relying on dense satellite image databases. We evaluate these models across two real-world environments: a university campus navigated by a Jackal robot and an urban downtown area navigated by a Tesla sedan. The VAE-Transformer achieves impressive precision, with median deviations of 2.29m (1.37% of environment size) and 4.45m (0.35% of environment size) respectively, outperforming both VAE-RNN and prior cross-view geo-localization approaches. Our comprehensive Localization Performance Characteristics (LPC) analysis demonstrates superior performance with the VAE-Transformer achieving an AUC of 0.777 compared to 0.295 for VIGOR 200 and 0.225 for TransGeo, establishing a new state-of-the-art in vision-based localization. In some scenarios, our vision-based system rivals commercial smartphone GPS accuracy (AUC of 0.797) while requiring 5x less GPU memory and delivering 3x faster inference than existing methods in cross-view geo-localization. These results demonstrate that models inspired by biological spatial navigation can effectively memorize complex, dynamic environments and provide precise localization with minimal computational resources.
- Open Access
Neural Mechanisms Linking Global Maps to First-Person Perspectives
Humans and many animals possess the remarkable ability to navigate environments by seamlessly switching between first-person perspectives (FPP) and global map perspectives (GMP). However, the neural mechanisms that underlie this transformation remain poorly understood. In this study, we developed a variational autoencoder (VAE) model, enhanced with recurrent neural networks (RNNs), to investigate the computational principles behind perspective transformations. Our results reveal that temporal sequence modeling is crucial for maintaining spatial continuity and improving transformation accuracy when switching between FPPs and GMPs. The model’s latent variables capture many representational forms seen in the distributed cognitive maps of the mammalian brain, such as head direction cells, place cells, corner cells, and border cells, but notably not grid cells, suggesting that perspective transformation engages multiple brain regions beyond the hippocampus and entorhinal cortex. Furthermore, our findings demonstrate that landmark encoding, particularly proximal environmental cues such as boundaries and objects, play a critical role in enabling successful perspective shifts, whereas distal cues are less influential. These insights on perspective linking provide a new computational framework for understanding spatial cognition and offer valuable directions for future animal and human studies, highlighting the significance of temporal sequences, distributed representations, and proximal cues in navigating complex environments. Significance Statement: Understanding how the brain transforms between different spatial perspectives is crucial for advancing our knowledge of spatial cognition and navigation. This study presents a novel computational approach that bridges the gap between neural recordings and behavior, offering insights into the underlying mechanisms of perspective transformation. Our findings suggest how the brain integrates temporal sequences, distributed representations, and environmental cues to maintain a coherent sense of space. By demonstrating the importance of proximal cues and temporal context, our computational model provides testable predictions for future neurophysiological studies in humans and animals.
Open AccessNeural Mechanisms Linking Global Maps to First-Person Perspectives
Humans and many animals possess the remarkable ability to navigate environments by seamlessly switching between first-person perspectives (FPP) and global map perspectives (GMP). However, the neural mechanisms that underlie this transformation remain poorly understood. In this study, we developed a variational autoencoder (VAE) model, enhanced with recurrent neural networks (RNNs), to investigate the computational principles behind perspective transformations. Our results reveal that temporal sequence modeling is crucial for maintaining spatial continuity and improving transformation accuracy when switching between FPPs and GMPs. The model’s latent variables capture many representational forms seen in the distributed cognitive maps of the mammalian brain, such as head direction cells, place cells, corner cells, and border cells, but notably not grid cells, suggesting that perspective transformation engages multiple brain regions beyond the hippocampus and entorhinal cortex. Furthermore, our findings demonstrate that landmark encoding, particularly proximal environmental cues such as boundaries and objects, play a critical role in enabling successful perspective shifts, whereas distal cues are less influential. These insights on perspective linking provide a new computational framework for understanding spatial cognition and offer valuable directions for future animal and human studies, highlighting the significance of temporal sequences, distributed representations, and proximal cues in navigating complex environments. Significance Statement: Understanding how the brain transforms between different spatial perspectives is crucial for advancing our knowledge of spatial cognition and navigation. This study presents a novel computational approach that bridges the gap between neural recordings and behavior, offering insights into the underlying mechanisms of perspective transformation. Our findings suggest how the brain integrates temporal sequences, distributed representations, and environmental cues to maintain a coherent sense of space. By demonstrating the importance of proximal cues and temporal context, our computational model provides testable predictions for future neurophysiological studies in humans and animals.
- Open Access
- 30 January 2025
- Published by IOP Publishing in Neuromorphic Computing and Engineering
- Vol. 5 (1) , 014003
An integrated toolbox for creating neuromorphic edge applications
Spiking neural networks (SNNs) and neuromorphic models are believed to be more efficient in general and have more biological realism than the activation functions typically used in deep neural networks, transformer models and generative AI. SNNs have local learning rules, are able to learn on small data sets, and can adapt through neuromodulation. Although research has shown their advantages, there are still few compelling practical applications, especially at the edge where sensors and actuators need to be processed in a timely fashion. One reason for this might be that SNNs are much more challenging to understand, build, and operate due to their intrinsic properties. For instance, the mathematical foundation involves differential equations rather than basic activation functions. To address these challenges, we have developed CARLsim++, an integrated toolbox that facilitates the creation of neuromorphic applications. It extends the highly efficient CARLsim open-source SNN simulator. CARLsim++ encapsulates the mathematical intrinsics and low-level C++ programming by providing a graphical user interface for users to easily create their SNNs and a means to configure sensors and actuators for robotics and other edge devices. These can be accurately simulated before deploying on physical devices. CARLsim++ can lead to rapid development of neuromorphic applications for simulation or edge processing. We introduce CARLsim++ with a closed loop robotic demonstration using neuromorphic computing.
- 30 January 2025
- Published by IOP Publishing in Neuromorphic Computing and Engineering
- Vol. 5 (1) , 014003
Open AccessAn integrated toolbox for creating neuromorphic edge applications
Spiking neural networks (SNNs) and neuromorphic models are believed to be more efficient in general and have more biological realism than the activation functions typically used in deep neural networks, transformer models and generative AI. SNNs have local learning rules, are able to learn on small data sets, and can adapt through neuromodulation. Although research has shown their advantages, there are still few compelling practical applications, especially at the edge where sensors and actuators need to be processed in a timely fashion. One reason for this might be that SNNs are much more challenging to understand, build, and operate due to their intrinsic properties. For instance, the mathematical foundation involves differential equations rather than basic activation functions. To address these challenges, we have developed CARLsim++, an integrated toolbox that facilitates the creation of neuromorphic applications. It extends the highly efficient CARLsim open-source SNN simulator. CARLsim++ encapsulates the mathematical intrinsics and low-level C++ programming by providing a graphical user interface for users to easily create their SNNs and a means to configure sensors and actuators for robotics and other edge devices. These can be accurately simulated before deploying on physical devices. CARLsim++ can lead to rapid development of neuromorphic applications for simulation or edge processing. We introduce CARLsim++ with a closed loop robotic demonstration using neuromorphic computing.