An Efficient Insect-inspired Approach for Visual Point-goal Navigation

arxiv: 2601.16806 · v3 · submitted 2026-01-23 · 💻 cs.AI · cs.RO

An Efficient Insect-inspired Approach for Visual Point-goal Navigation

Yihe Lu , Barbara Webb This is my paper

Pith reviewed 2026-05-16 12:01 UTC · model grok-4.3

classification 💻 cs.AI cs.RO

keywords insect-inspired navigationpoint-goal navigationassociative learningpath integrationvisual navigationefficient modelsHabitat benchmarkbrain structures

0 comments p. Extension

The pith

An insect-inspired model using brain structures for learning and path integration performs point-goal navigation as well as advanced AI systems at far lower computational cost.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper develops a model that combines simplified versions of insect brain areas involved in associative learning and path integration to address visual point-goal navigation. It maps this to the Habitat benchmark task, where an agent must reach a goal location using visual input while avoiding obstacles. The resulting system matches the success rates of recent high-performance models while requiring many orders of magnitude less computation and holds up when tested with added perturbations in more realistic simulated settings. A sympathetic reader would care because navigation under visual constraints is fundamental for autonomous agents, and a low-cost biological approach could make such capabilities practical on limited hardware.

Core claim

The central claim is that a model formed by integrating an abstracted associative learning circuit modeled on the insect mushroom body with a path integration circuit modeled on the central complex can solve the Habitat point-goal navigation task. This combination allows the agent to discover, learn, and refine visually guided routes around obstacles in a manner analogous to insect foraging between food sites and the nest. When evaluated on the standard benchmark, the approach reaches performance levels comparable to recent state-of-the-art methods while incurring far lower computational expense, and it continues to function reliably under environmental perturbations in extended simulation.

What carries the argument

The central mechanism is the integrated model of insect associative learning and path integration, which together enable the agent to store and recall visual routes while maintaining an internal estimate of displacement to the goal.

Load-bearing premise

That abstracted models of insect brain structures for associative learning and path integration can be directly combined and applied to the Habitat point-goal navigation task while preserving the claimed efficiency and robustness.

What would settle it

Running the described model on the standard Habitat point-goal navigation benchmark and measuring success rate, SPL score, and computational cost; the claim would be falsified if success metrics fall substantially below those of recent state-of-the-art models or if resource usage does not remain orders of magnitude lower.

Figures

Figures reproduced from arXiv: 2601.16806 by Barbara Webb, Yihe Lu.

**Figure 1.** Figure 1: Overview of our insect-inspired model. (A) The closed-loop learning and control of our model embodied in a virtual robot. The model is primarily composed of two insect brain-like modules, a mushroom body (MB) and a central complex (CX), They use three types of (preprocessed) data as inputs, odometry (I), collision (II), and vision (III) to calculate a desired target direction, which determines control comm… view at source ↗

**Figure 2.** Figure 2: Model performance (SR (square) and SPL (cross), defined in §2.2.4) vs number of training frames. The [PITH_FULL_IMAGE:figures/full_fig_p008_2.png] view at source ↗

**Figure 3.** Figure 3: Performance of insect-inspired models in [PITH_FULL_IMAGE:figures/full_fig_p009_3.png] view at source ↗

**Figure 4.** Figure 4: Summary of model collisions. Note the scales are different for the ablated vs full models. [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗

**Figure 5.** Figure 5: Performance of insect-inspired models with different memory consolidation mechanisms. Full-learnt: our [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗

**Figure 6.** Figure 6: An example experiment with strong motor perturbation, [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗

**Figure 7.** Figure 7: The effect of motor noise on performance. As before, blue is the full model and red the odometry-collision [PITH_FULL_IMAGE:figures/full_fig_p012_7.png] view at source ↗

**Figure 8.** Figure 8: The effect of motor bias on performance. As before, blue is the full model and red the odometry-collision [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗

**Figure 9.** Figure 9: Two hypothetical point-goal navigation episodes. The (black) obstacles are symmetric in the beeline [PITH_FULL_IMAGE:figures/full_fig_p015_9.png] view at source ↗

**Figure 10.** Figure 10: Zoom-in plots of the area enclosed in the rectangle with dashed boundaries in Fig. 6. The black circles here [PITH_FULL_IMAGE:figures/full_fig_p021_10.png] view at source ↗

**Figure 11.** Figure 11: The effect of motor noise on performance, across trials within an episode. Blue is the full model, red is [PITH_FULL_IMAGE:figures/full_fig_p022_11.png] view at source ↗

**Figure 12.** Figure 12: The effect of motor bias on performance, across trials within an episode. Blue is the full model, red is [PITH_FULL_IMAGE:figures/full_fig_p022_12.png] view at source ↗

read the original abstract

In this work we develop a novel insect-inspired model for visual point-goal navigation. This combines abstracted models of two insect brain structures that have been implicated, respectively, in associative learning and path integration. We draw an analogy between the formal benchmark of the Habitat point-goal navigation task and the ability of insects to discover, learn, and refine visually guided paths around obstacles between a discovered food location and their nest. We demonstrate that the simple insect-inspired model exhibits performance comparable to recent state-of-the-art models at many orders of magnitude less computational cost. Testing in a more realistic simulated environment shows the approach is robust to perturbations.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows a parameter-free insect-inspired model reaching Habitat-comparable point-goal navigation at far lower cost, with the main value in the clean mapping of established neuroscience abstractions to the benchmark.

read the letter

Colleague, the core result here is that abstracted models of insect structures for associative learning and path integration, when combined, produce navigation success rates and SPL scores close to recent learned methods on the Habitat point-goal task while using orders of magnitude less compute. They also report robustness under perturbations in a more realistic simulator. That efficiency angle is the part worth paying attention to if the numbers check out. What the work actually does is take two established insect neuroscience components, treat the Habitat setup as analogous to an insect discovering and refining a path between food and nest, and implement the whole thing without free parameters. The mapping from brain structures to observation and action spaces appears direct enough that the construction stays testable rather than circular. Credit is due for keeping the model simple and for running the extra robustness check instead of stopping at the standard benchmark. The soft spot is that the abstract gives no concrete metrics, baselines, or error analysis, so the comparability claim cannot be judged from the summary alone. If the full paper supplies the exact translation rules and the raw success rates against the cited SOTA models, the efficiency result becomes verifiable; without those numbers the central claim stays plausible but unconfirmed. Minor additional concern is whether the insect abstractions were chosen because they fit the task or because they are the right biological priors; the paper does not appear to overclaim generality beyond the tested setting. This is the kind of paper that matters for groups working on low-resource visual navigation or on tighter neuroscience-AI links. Readers who care about reproducible, parameter-light methods on standard benchmarks will find it useful; those looking for broad theoretical advances will not. The thinking is coherent on its own terms and the construction is falsifiable, so the work deserves a serious referee to examine the implementation details and the quantitative results. I would send it out for review rather than desk reject.

Referee Report

2 major / 0 minor

Summary. The manuscript presents a novel insect-inspired model for visual point-goal navigation that integrates abstracted models of insect brain structures implicated in associative learning and path integration. It draws an explicit analogy between the Habitat point-goal navigation benchmark and insect foraging behavior, claiming that the resulting simple model achieves performance comparable to recent state-of-the-art approaches at many orders of magnitude lower computational cost while remaining robust under perturbations in more realistic simulated environments.

Significance. If the quantitative claims hold, the work would be significant for showing that a parameter-free, biologically abstracted construction can match the navigation performance of complex learned models on the Habitat benchmark. The zero free parameters, direct mapping from insect-inspired rules to observation and action spaces, and explicit testability of the efficiency and robustness statements are notable strengths that distinguish this from typical fitted neural approaches.

major comments (2)

Abstract: the central claim that the model 'exhibits performance comparable to recent state-of-the-art models' is stated without any quantitative metrics (success rate, SPL, runtime, or baseline tables), making the comparability assertion impossible to evaluate from the provided text.
Abstract and results description: the robustness claim ('robust to perturbations') lacks any specification of the perturbation types, the quantitative change in success/SPL, or error bars, which is load-bearing for the robustness part of the central claim.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for their constructive comments and positive assessment of the work's significance. We address each major comment below and will revise the manuscript to improve clarity and evaluability of the claims.

read point-by-point responses

Referee: Abstract: the central claim that the model 'exhibits performance comparable to recent state-of-the-art models' is stated without any quantitative metrics (success rate, SPL, runtime, or baseline tables), making the comparability assertion impossible to evaluate from the provided text.

Authors: We agree that the abstract should include quantitative metrics to allow direct evaluation. The full manuscript reports these comparisons (success rates, SPL scores, and runtime) against state-of-the-art baselines in the results section, along with the orders-of-magnitude computational savings. We will revise the abstract to incorporate the key quantitative results from our experiments. revision: yes
Referee: Abstract and results description: the robustness claim ('robust to perturbations') lacks any specification of the perturbation types, the quantitative change in success/SPL, or error bars, which is load-bearing for the robustness part of the central claim.

Authors: We acknowledge the need for greater specificity. The manuscript tests robustness under perturbations in more realistic simulated environments and reports performance changes, but these details are not summarized in the abstract or results overview. We will revise the abstract and results description to explicitly list the perturbation types, quantify the changes in success rate and SPL, and include error bars. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The derivation combines established abstractions of insect associative learning and path integration (drawn from external neuroscience literature) and directly maps them onto the Habitat observation/action spaces. No parameter is fitted to the target benchmark data and then relabeled as a prediction; no quantity is defined in terms of itself; no uniqueness theorem or ansatz is imported solely via self-citation to force the result; and the efficiency/robustness claims are evaluated on an external simulator rather than being tautological. The construction is therefore self-contained against the stated benchmark.

Axiom & Free-Parameter Ledger

0 free parameters · 1 axioms · 0 invented entities

The central claim rests on the assumption that insect navigation mechanisms can be abstracted into simple computational models that transfer effectively to simulated visual navigation benchmarks.

axioms (1)

domain assumption Insect brain structures implicated in associative learning and path integration can be abstracted into computational models that solve navigation tasks.
Core premise drawn from neuroscience literature to justify the model construction.

pith-pipeline@v0.9.0 · 5387 in / 1096 out tokens · 39283 ms · 2026-05-16T12:01:00.671343+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The MB architecture... y(t)=g(W0·x(t)), z(t)=W1(t)·y(t)/k, ΔWji=−α·yi(t−τe)·Wji(t−τe); CX: Δσ=θ−σ+ϕ, ϕ=π(zL−zR)
IndisputableMonolith/Foundation/ArithmeticFromLogic.lean LogicNat induction and orbit embedding unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Memory consolidation... ΔW1=ΔWLTM+ΔWITM+ΔWSTM; selective consolidation based on SPL improvement

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.

Reference graph

Works this paper leans on

73 extracted references · 73 canonical work pages · 3 internal anchors

[1]

Habitat: a platform for embodied AI research

Manolis Savva, Abhishek Kadian, Oleksandr Maksymets, Yili Zhao, Erik Wijmans, Bhavana Jain, Julian Straub, Jia Liu, Vladlen Koltun, Jitendra Malik, et al. Habitat: a platform for embodied AI research. InProceedings of the IEEE/CVF international conference on computer vision, pages 9339–9347, 2019

work page 2019
[2]

Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI

Santhosh K Ramakrishnan, Aaron Gokaslan, Erik Wijmans, Oleksandr Maksymets, Alex Clegg, John Turner, Eric Undersander, Wojciech Galuba, Andrew Westbury, Angel X Chang, et al. Habitat-Matterport 3D dataset (HM3D): 1000 large-scale 3D environments for embodied AI.arXiv preprint arXiv:2109.08238, 2021

work page internal anchor Pith review Pith/arXiv arXiv 2021
[3]

Decentralized distributed PPO: solving pointgoal navigation

Erik Wijmans, Abhishek Kadian, Ari Morcos, Stefan Lee, Irfan Essa, Devi Parikh, Manolis Savva, and Dhruv Batra. DD-PPO: learning near-perfect pointgoal navigators from 2.5 billion frames.arXiv preprint arXiv:1911.00357, 2019

work page arXiv 1911
[4]

Path integration in desert ants,Cataglyphis fortis.Proceedings of the National Academy of Sciences, 85(14):5287–5290, 1988

Martin Müller and Rüdiger Wehner. Path integration in desert ants,Cataglyphis fortis.Proceedings of the National Academy of Sciences, 85(14):5287–5290, 1988. 17 APREPRINT- JANUARY26, 2026

work page 1988
[5]

Visual landmarks and route following in desert ants.Journal of Comparative Physiology A, 170:435–442, 1992

Thomas S Collett, Elisabeth Dillmann, A Giger, and Rüdiger Wehner. Visual landmarks and route following in desert ants.Journal of Comparative Physiology A, 170:435–442, 1992

work page 1992
[6]

Views, landmarks, and routes: how do desert ants negotiate an obstacle course?Journal of Comparative Physiology A, 197:167–179, 2011

Antoine Wystrach, Sebastian Schwarz, Patrick Schultheiss, Guy Beugnon, and Ken Cheng. Views, landmarks, and routes: how do desert ants negotiate an obstacle course?Journal of Comparative Physiology A, 197:167–179, 2011

work page 2011
[7]

Visual learning, route formation and the choreography of looking back in desert ants,Melophorus bagoti.Animal Behaviour, 222:123125, 2025

Cody A Freas and Ken Cheng. Visual learning, route formation and the choreography of looking back in desert ants,Melophorus bagoti.Animal Behaviour, 222:123125, 2025

work page 2025
[8]

The neuronal architecture of the mushroom body provides a logic for associative learning.eLife, 3:e04577, 2014

Yoshinori Aso, Daisuke Hattori, Yang Yu, Rebecca M Johnston, Nirmala A Iyer, Teri-TB Ngo, Heather Dionne, LF Abbott, Richard Axel, Hiromu Tanimoto, et al. The neuronal architecture of the mushroom body provides a logic for associative learning.eLife, 3:e04577, 2014

work page 2014
[9]

Beyond prediction error: 25 years of modeling the associations formed in the insect mushroom body.Learning & Memory, 31(5):a053824, 2024

Barbara Webb. Beyond prediction error: 25 years of modeling the associations formed in the insect mushroom body.Learning & Memory, 31(5):a053824, 2024

work page 2024
[10]

Representations of novelty and familiarity in a mushroom body compartment.Cell, 169(5):956–969, 2017

Daisuke Hattori, Yoshinori Aso, Kurtis J Swartz, Gerald M Rubin, L F Abbott, and Richard Axel. Representations of novelty and familiarity in a mushroom body compartment.Cell, 169(5):956–969, 2017

work page 2017
[11]

An anatomically constrained model for path integration in the bee brain.Current Biology, 27(20):3069–3085, 2017

Thomas Stone, Barbara Webb, Andrea Adden, Nicolai Ben Weddig, Anna Honkanen, Rachel Templin, William Wcislo, Luca Scimeca, Eric Warrant, and Stanley Heinze. An anatomically constrained model for path integration in the bee brain.Current Biology, 27(20):3069–3085, 2017

work page 2017
[12]

The internal maps of insects.Journal of Experimental Biology, 222(Suppl_1):jeb188094, 2019

Barbara Webb. The internal maps of insects.Journal of Experimental Biology, 222(Suppl_1):jeb188094, 2019

work page 2019
[13]

A connectome of theDrosophilacentral complex reveals network motifs suitable for flexible navigation and context-dependent action selection.eLife, 10, 2021

Brad K Hulse, Hannah Haberkern, Romain Franconville, Daniel Turner-Evans, Shin-ya Takemura, Tanya Wolff, Marcella Noorman, Marisa Dreher, Chuntao Dan, Ruchi Parekh, et al. A connectome of theDrosophilacentral complex reveals network motifs suitable for flexible navigation and context-dependent action selection.eLife, 10, 2021

work page 2021
[14]

The central complex as a potential substrate for vector based navigation.Frontiers in Psychology, 10:690, 2019

Florent Le Moël, Thomas Stone, Mathieu Lihoreau, Antoine Wystrach, and Barbara Webb. The central complex as a potential substrate for vector based navigation.Frontiers in Psychology, 10:690, 2019

work page 2019
[15]

A neurocomputational model of goal-directed navigation in insect-inspired artificial agents.Frontiers in Neurorobotics, 11:20, 2017

Dennis Goldschmidt, Poramate Manoonpong, and Sakyasingha Dasgupta. A neurocomputational model of goal-directed navigation in insect-inspired artificial agents.Frontiers in Neurorobotics, 11:20, 2017

work page 2017
[16]

A decentralised neural model explaining optimal integration of navigational strategies in insects.eLife, 9:e54026, 2020

Xuelong Sun, Shigang Yue, and Michael Mangan. A decentralised neural model explaining optimal integration of navigational strategies in insects.eLife, 9:e54026, 2020

work page 2020
[17]

Emergent spatial goals in an integrative model of the insect central complex.PLOS Computational Biology, 19(12):e1011480, 2023

Roman Goulard, Stanley Heinze, and Barbara Webb. Emergent spatial goals in an integrative model of the insect central complex.PLOS Computational Biology, 19(12):e1011480, 2023

work page 2023
[18]

Neurons from pre-motor areas to the mushroom bodies can orchestrate latent visual learning in navigating insects.bioRxiv, 2023

Antoine Wystrach. Neurons from pre-motor areas to the mushroom bodies can orchestrate latent visual learning in navigating insects.bioRxiv, 2023

work page 2023
[19]

Using the neural circuit of the insect central complex for path integration on a micro aerial vehicle

Jan Stankiewicz and Barbara Webb. Using the neural circuit of the insect central complex for path integration on a micro aerial vehicle. InConference on Biomimetic and Biohybrid Systems, pages 325–337. Springer, 2020

work page 2020
[20]

A model of ant route navigation driven by scene familiarity.PLOS Computational Biology, 8(1):e1002336, 2012

Bart Baddeley, Paul Graham, Philip Husbands, and Andrew Philippides. A model of ant route navigation driven by scene familiarity.PLOS Computational Biology, 8(1):e1002336, 2012

work page 2012
[21]

An insect-inspired model facilitating autonomous navigation by incorporating goal approaching and collision avoidance.Neural Networks, 165:106–118, 2023

Xuelong Sun, Qinbing Fu, Jigen Peng, and Shigang Yue. An insect-inspired model facilitating autonomous navigation by incorporating goal approaching and collision avoidance.Neural Networks, 165:106–118, 2023

work page 2023
[22]

How might ants use panoramic views for route navigation?Journal of Experimental Biology, 214(3):445–451, 2011

Andrew Philippides, Bart Baddeley, Ken Cheng, and Paul Graham. How might ants use panoramic views for route navigation?Journal of Experimental Biology, 214(3):445–451, 2011

work page 2011
[23]

The visual compass: Performance and limitations of an appearance-based method.Journal of Field Robotics, 23(10):913–941, 2006

Frédéric Labrosse. The visual compass: Performance and limitations of an appearance-based method.Journal of Field Robotics, 23(10):913–941, 2006

work page 2006
[24]

Using an insect mush- room body circuit to encode route memory in complex natural environments.PLOS Computational Biology, 12(2):e1004683, 2016

Paul Ardin, Fei Peng, Michael Mangan, Konstantinos Lagogiannis, and Barbara Webb. Using an insect mush- room body circuit to encode route memory in complex natural environments.PLOS Computational Biology, 12(2):e1004683, 2016

work page 2016
[25]

Route following without scanning

Aleksandar Kodzhabashev and Michael Mangan. Route following without scanning. In Stuart P. Wilson, Paul F.M.J. Verschure, Anna Mura, and Tony J. Prescott, editors,Biomimetic and Biohybrid Systems., pages 199–210, Cham, 2015. Springer International Publishing

work page 2015
[26]

Familiarity-taxis: a bilateral approach to view-based snapshot navigation.Adaptive Behavior, 32(5):407–420, October 2024

Fabian Steinbeck, Efstathios Kagioulis, Alex Dewar, Andrew Philippides, Thomas Nowotny, and Paul Graham. Familiarity-taxis: a bilateral approach to view-based snapshot navigation.Adaptive Behavior, 32(5):407–420, October 2024. 18 APREPRINT- JANUARY26, 2026

work page 2024
[27]

Insect-inspired embodied visual route following

Yihe Lu, Jiahao Cen, Rana Maroun Alkhoury, and Barbara Webb. Insect-inspired embodied visual route following. Journal of Bionic Engineering, 2025

work page 2025
[28]

Insect visual homing strategies in a robot with analog processing.Biological Cybernetics, 83(3):231– 243, 2000

Ralf Möller. Insect visual homing strategies in a robot with analog processing.Biological Cybernetics, 83(3):231– 243, 2000

work page 2000
[29]

Learning and time-dependent cue choice in the desert ant,Melophorus bagoti

Cody A Freas and Ken Cheng. Learning and time-dependent cue choice in the desert ant,Melophorus bagoti. Ethology, 123(8):503–515, 2017

work page 2017
[30]

Terrestrial cue learning and retention during the outbound and inbound foraging trip in the desert ant, cataglyphis velox.Journal of Comparative Physiology A, 205(2):177–189, 2019

Cody A Freas and Marcia L Spetch. Terrestrial cue learning and retention during the outbound and inbound foraging trip in the desert ant, cataglyphis velox.Journal of Comparative Physiology A, 205(2):177–189, 2019

work page 2019
[31]

Latent learning without map-like representation of space in navigating ants.bioRxiv, pages 2024–08, 2024

Leo Clement, Sebastian Schwarz, and Antoine Wystrach. Latent learning without map-like representation of space in navigating ants.bioRxiv, pages 2024–08, 2024

work page 2024
[32]

iGibson 2.0: object-centric simulation for robot learning of everyday household tasks

Chengshu Li, Fei Xia, Roberto Martín-Martín, Michael Lingelbach, Sanjana Srivastava, Bokui Shen, Kent Elliott Vainio, Cem Gokmen, Gokul Dharan, Tanish Jain, Andrey Kurenkov, Karen Liu, Hyowon Gweon, Jiajun Wu, Li Fei-Fei, and Silvio Savarese. iGibson 2.0: object-centric simulation for robot learning of everyday household tasks. InProceedings of the 5th Co...

work page 2022
[33]

Theory of edge detection.Proceedings of the Royal Society of London

David Marr and Ellen Hildreth. Theory of edge detection.Proceedings of the Royal Society of London. Series B. Biological Sciences, 207(1167):187–217, 1980

work page 1980
[34]

The properties of the visual system in the Australian desert antMelophorus bagoti.Arthropod Structure & Development, 40(2):128–134, 2011

Sebastian Schwarz, Ajay Narendra, and Jochen Zeil. The properties of the visual system in the Australian desert antMelophorus bagoti.Arthropod Structure & Development, 40(2):128–134, 2011

work page 2011
[35]

Predicting visual function by interpreting a neuronal wiring diagram.Nature, 634(8032):113– 123, 2024

H Sebastian Seung. Predicting visual function by interpreting a neuronal wiring diagram.Nature, 634(8032):113– 123, 2024

work page 2024
[36]

Antoine Wystrach, Alex Dewar, Andrew Philippides, and Paul Graham. How do field of view and resolution affect the information content of panoramic scenes for visual navigation? A computational investigation.Journal of Comparative Physiology A, 202:87–95, 2016

work page 2016
[37]

Route-centric ant-inspired memories enable panoramic route-following in a car-like robot.Nature Communications, 16(1):8328, 2025

Gabriel G Gattaux, Antoine Wystrach, Julien R Serres, and Franck Ruffier. Route-centric ant-inspired memories enable panoramic route-following in a car-like robot.Nature Communications, 16(1):8328, 2025

work page 2025
[38]

Isotropic-sequence-order learning in a closed-loop behavioural system

Bernd Porr and Florentin Wörgötter. Isotropic-sequence-order learning in a closed-loop behavioural system. Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences, 361(1811):2225–2244, 2003

work page 2003
[39]

Integration of parallel opposing memories underlies memory extinction.Cell, 175(3):709–722, 2018

Johannes Felsenberg, Pedro F Jacob, Thomas Walker, Oliver Barnstedt, Amelia J Edmondson-Stait, Markus W Pleijzier, Nils Otto, Philipp Schlegel, Nadiya Sharifi, Emmanuel Perisse, et al. Integration of parallel opposing memories underlies memory extinction.Cell, 175(3):709–722, 2018

work page 2018
[40]

A neural data structure for novelty detection.Proceedings of the National Academy of Sciences, 115(51):13093–13098, 2018

Sanjoy Dasgupta, Timothy C Sheehan, Charles F Stevens, and Saket Navlakha. A neural data structure for novelty detection.Proceedings of the National Academy of Sciences, 115(51):13093–13098, 2018

work page 2018
[41]

Traces of drosophila memory.Neuron, 70(1):8–19, 2011

Ronald L Davis. Traces of drosophila memory.Neuron, 70(1):8–19, 2011

work page 2011
[42]

An incentive circuit for memory dynamics in the mushroom body ofDrosophila melanogaster.eLife, 11:e75611, 2022

Evripidis Gkanias, Li Yan McCurdy, Michael N Nitabach, and Barbara Webb. An incentive circuit for memory dynamics in the mushroom body ofDrosophila melanogaster.eLife, 11:e75611, 2022

work page 2022
[43]

iGibson 1.0: a simulation environment for interactive tasks in large realistic scenes

Bokui Shen, Fei Xia, Chengshu Li, Roberto Martín-Martín, Linxi Fan, Guanzhi Wang, Claudia Pérez-D’Arpino, Shyamal Buch, Sanjana Srivastava, Lyne Tchapmi, et al. iGibson 1.0: a simulation environment for interactive tasks in large realistic scenes. In2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 7520–7527. IEEE, 2021

work page 2021
[44]

Zamir, Zhi-Yang He, Alexander Sax, Jitendra Malik, and Silvio Savarese

Fei Xia, Amir R. Zamir, Zhi-Yang He, Alexander Sax, Jitendra Malik, and Silvio Savarese. Gibson Env: real-world perception for embodied agents. InComputer Vision and Pattern Recognition (CVPR), 2018 IEEE Conference on. IEEE, 2018

work page 2018
[45]

Benchmarking Classic and Learned Navigation in Complex 3D Environments

Dmytro Mishkin, Alexey Dosovitskiy, and Vladlen Koltun. Benchmarking classic and learned navigation in complex 3D environments.arXiv preprint arXiv:1901.10915, 2019

work page internal anchor Pith review Pith/arXiv arXiv 1901
[46]

The surprising effectiveness of visual odometry techniques for embodied pointgoal navigation

Xiaoming Zhao, Harsh Agrawal, Dhruv Batra, and Alexander Schwing. The surprising effectiveness of visual odometry techniques for embodied pointgoal navigation. In2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 16107–16116, 2021

work page 2021
[47]

Is mapping necessary for realistic pointgoal navigation? In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17211–17220, 2022

Ruslan Partsey, Erik Wijmans, Naoki Yokoyama, Oles Dobosevych, Dhruv Batra, and Oleksandr Maksymets. Is mapping necessary for realistic pointgoal navigation? In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17211–17220, 2022. 19 APREPRINT- JANUARY26, 2026

work page 2022
[48]

Sim2real predictivity: does evaluation in simulation predict real-world performance?IEEE Robotics and Automation Letters, 5(4):6670–6677, 2020

Abhishek Kadian, Joanne Truong, Aaron Gokaslan, Alexander Clegg, Erik Wijmans, Stefan Lee, Manolis Savva, Sonia Chernova, and Dhruv Batra. Sim2real predictivity: does evaluation in simulation predict real-world performance?IEEE Robotics and Automation Letters, 5(4):6670–6677, 2020

work page 2020
[49]

MINOS: Multimodal Indoor Simulator for Navigation in Complex Environments

Manolis Savva, Angel X. Chang, Alexey Dosovitskiy, Thomas Funkhouser, and Vladlen Koltun. MINOS: multimodal indoor simulator for navigation in complex environments.arXiv:1712.03931, 2017

work page internal anchor Pith review Pith/arXiv arXiv 2017
[50]

Cognitive architecture of a mini-brain: the honeybee.Trends in Cognitive Sciences, 5(2):62–71, 2001

Randolf Menzel and Martin Giurfa. Cognitive architecture of a mini-brain: the honeybee.Trends in Cognitive Sciences, 5(2):62–71, 2001

work page 2001
[51]

Allometric analysis of brain cell num- ber in Hymenoptera suggests ant brains diverge from general trends.Proceedings of the Royal Society B, 288(1947):20210199, 2021

R Keating Godfrey, Mira Swartzlander, and Wulfila Gronenberg. Allometric analysis of brain cell num- ber in Hymenoptera suggests ant brains diverge from general trends.Proceedings of the Royal Society B, 288(1947):20210199, 2021

work page 1947
[52]

A lateralised design for the interaction of visual memories and heading representations in navigating ants.bioRxiv, 2020

Antoine Wystrach, Florent Le Moël, Leo Clement, and Sebastian Schwarz. A lateralised design for the interaction of visual memories and heading representations in navigating ants.bioRxiv, 2020

work page 2020
[53]

Obstacle avoidance in a crowd with a plastic recurrent mushroom body model.IEEE Access, 2025

Liying Tao, Zonglin Yang, Gaoming Li, and Delong Shang. Obstacle avoidance in a crowd with a plastic recurrent mushroom body model.IEEE Access, 2025

work page 2025
[54]

A comparative study of bug algorithms for robot navigation.Robotics and Autonomous Systems, 121:103261, 2019

Kimberly N McGuire, Guido C H E de Croon, and Karl Tuyls. A comparative study of bug algorithms for robot navigation.Robotics and Autonomous Systems, 121:103261, 2019

work page 2019
[55]

Lumelsky and A

V . Lumelsky and A. Stepanov. Dynamic path planning for a mobile automaton with limited information on the environment.IEEE Transactions on Automatic Control, 31(11):1058–1063, 1986

work page 1986
[56]

Learning with reinforcement prediction errors in a model of theDrosophilamushroom body.Nature Communications, 12(1):2569, 2021

James EM Bennett, Andrew Philippides, and Thomas Nowotny. Learning with reinforcement prediction errors in a model of theDrosophilamushroom body.Nature Communications, 12(1):2569, 2021

work page 2021
[57]

Prediction error drives associative learning and conditioned behavior in a spiking model of Drosophila larva

Anna-Maria Jürgensen, Panagiotis Sakagiannis, Michael Schleyer, Bertram Gerber, and Martin Paul Nawrot. Prediction error drives associative learning and conditioned behavior in a spiking model of Drosophila larva. iScience, 27(1), 2024

work page 2024
[58]

Stephan Lochner, Daniel Honerkamp, Abhinav Valada, and Andrew D Straw. Reinforcement learning as a robotics- inspired framework for insect navigation: from spatial representations to neural implementation.Frontiers in Computational Neuroscience, 18:1460006, 2024

work page 2024
[59]

Sleep deprivation results in diverse patterns of synaptic scaling across theDrosophilamushroom bodies.Current Biology, 31(15):3248–3261, 2021

Jacqueline T Weiss and Jeffrey M Donlea. Sleep deprivation results in diverse patterns of synaptic scaling across theDrosophilamushroom bodies.Current Biology, 31(15):3248–3261, 2021

work page 2021
[60]

Catastrophic forgetting in connectionist networks.Trends in Cognitive Sciences, 3(4):128–135, 1999

Robert M French. Catastrophic forgetting in connectionist networks.Trends in Cognitive Sciences, 3(4):128–135, 1999

work page 1999
[61]

Processing of expected and unexpected events during conditioning and attention: a psy- chophysiological theory.Psychological Review, 89(5):529, 1982

Stephen Grossberg. Processing of expected and unexpected events during conditioning and attention: a psy- chophysiological theory.Psychological Review, 89(5):529, 1982

work page 1982
[62]

Reducing catastrophic forgetting with associative learning: a lesson from fruit flies.Neural Computation, 35(11):1797–1819, 2023

Yang Shen, Sanjoy Dasgupta, and Saket Navlakha. Reducing catastrophic forgetting with associative learning: a lesson from fruit flies.Neural Computation, 35(11):1797–1819, 2023

work page 2023
[63]

Large scale homing in honeybees.PLOS One, 6(5):e19669, 2011

Mario Pahl, Hong Zhu, Jürgen Tautz, and Shaowu Zhang. Large scale homing in honeybees.PLOS One, 6(5):e19669, 2011

work page 2011
[64]

Ants might use different view-matching strategies on and off the route.Journal of Experimental Biology, 215(1):44–55, 2012

Antoine Wystrach, Guy Beugnon, and Ken Cheng. Ants might use different view-matching strategies on and off the route.Journal of Experimental Biology, 215(1):44–55, 2012

work page 2012
[65]

Mushroom body output neurons encode valence and guide memory-based action selection in drosophila.eLife, 3:e04580, 2014

Yoshinori Aso, Divya Sitaraman, Toshiharu Ichinose, Karla R Kaun, Katrin V ogt, Ghislain Belliart-Guérin, Pierre-Yves Plaçais, Alice A Robie, Nobuhiro Yamagata, Christopher Schnaitmann, et al. Mushroom body output neurons encode valence and guide memory-based action selection in drosophila.eLife, 3:e04580, 2014

work page 2014
[66]

Opponent processes in visual memories: A model of attraction and repulsion in navigating insects’ mushroom bodies.PLOS Computational Biology, 16(2):e1007631, 2020

Florent Le Möel and Antoine Wystrach. Opponent processes in visual memories: A model of attraction and repulsion in navigating insects’ mushroom bodies.PLOS Computational Biology, 16(2):e1007631, 2020

work page 2020
[67]

A comparative study of reinforcement learning and insect-inspired visual navigation methods

Xiaoting Zhong, Xuelong Sun, and Haiyang Li. A comparative study of reinforcement learning and insect-inspired visual navigation methods. InConference on Biomimetic and Biohybrid Systems, pages 163–175. Springer, 2025. 20 APREPRINT- JANUARY26, 2026 5 Appendix 5.1 Model Details See Table 3. Table 3: More details of the recent point-goal navigation models i...

work page 2025
[68]

SunCG, Matterport3D ORB-SLAM2, D* Lite — CNN

work page
[69]

Gibson 4+, Matterport3D PPO/inherited from [45] SR++, geodesic-, frame- CNN

work page
[70]

Gibson 2+, Matterport3D DDPPO SPL++, geodesic-, frame- SE-ResNeXt50

work page
[71]

Gibson 4+ inherited from [3] inherited from [3] ResNet18

work page
[72]

Gibson 0+ inherited from [3] inherited from [3] ResNet50

work page
[73]

HM3D inherited from [3] inherited from [3] ResNet50 5.2 Supplementary Figures See Fig. 10-12. Figure 10: Zoom-in plots of the area enclosed in the rectangle with dashed boundaries in Fig. 6. The black circles here denote where collisions were detected. 21 APREPRINT- JANUARY26, 2026 Figure 11: The effect of motor noise on performance, across trials within ...

work page 2026

[1] [1]

Habitat: a platform for embodied AI research

Manolis Savva, Abhishek Kadian, Oleksandr Maksymets, Yili Zhao, Erik Wijmans, Bhavana Jain, Julian Straub, Jia Liu, Vladlen Koltun, Jitendra Malik, et al. Habitat: a platform for embodied AI research. InProceedings of the IEEE/CVF international conference on computer vision, pages 9339–9347, 2019

work page 2019

[2] [2]

Habitat-Matterport 3D Dataset (HM3D): 1000 Large-scale 3D Environments for Embodied AI

Santhosh K Ramakrishnan, Aaron Gokaslan, Erik Wijmans, Oleksandr Maksymets, Alex Clegg, John Turner, Eric Undersander, Wojciech Galuba, Andrew Westbury, Angel X Chang, et al. Habitat-Matterport 3D dataset (HM3D): 1000 large-scale 3D environments for embodied AI.arXiv preprint arXiv:2109.08238, 2021

work page internal anchor Pith review Pith/arXiv arXiv 2021

[3] [3]

Decentralized distributed PPO: solving pointgoal navigation

Erik Wijmans, Abhishek Kadian, Ari Morcos, Stefan Lee, Irfan Essa, Devi Parikh, Manolis Savva, and Dhruv Batra. DD-PPO: learning near-perfect pointgoal navigators from 2.5 billion frames.arXiv preprint arXiv:1911.00357, 2019

work page arXiv 1911

[4] [4]

Path integration in desert ants,Cataglyphis fortis.Proceedings of the National Academy of Sciences, 85(14):5287–5290, 1988

Martin Müller and Rüdiger Wehner. Path integration in desert ants,Cataglyphis fortis.Proceedings of the National Academy of Sciences, 85(14):5287–5290, 1988. 17 APREPRINT- JANUARY26, 2026

work page 1988

[5] [5]

Visual landmarks and route following in desert ants.Journal of Comparative Physiology A, 170:435–442, 1992

Thomas S Collett, Elisabeth Dillmann, A Giger, and Rüdiger Wehner. Visual landmarks and route following in desert ants.Journal of Comparative Physiology A, 170:435–442, 1992

work page 1992

[6] [6]

Views, landmarks, and routes: how do desert ants negotiate an obstacle course?Journal of Comparative Physiology A, 197:167–179, 2011

Antoine Wystrach, Sebastian Schwarz, Patrick Schultheiss, Guy Beugnon, and Ken Cheng. Views, landmarks, and routes: how do desert ants negotiate an obstacle course?Journal of Comparative Physiology A, 197:167–179, 2011

work page 2011

[7] [7]

Visual learning, route formation and the choreography of looking back in desert ants,Melophorus bagoti.Animal Behaviour, 222:123125, 2025

Cody A Freas and Ken Cheng. Visual learning, route formation and the choreography of looking back in desert ants,Melophorus bagoti.Animal Behaviour, 222:123125, 2025

work page 2025

[8] [8]

The neuronal architecture of the mushroom body provides a logic for associative learning.eLife, 3:e04577, 2014

Yoshinori Aso, Daisuke Hattori, Yang Yu, Rebecca M Johnston, Nirmala A Iyer, Teri-TB Ngo, Heather Dionne, LF Abbott, Richard Axel, Hiromu Tanimoto, et al. The neuronal architecture of the mushroom body provides a logic for associative learning.eLife, 3:e04577, 2014

work page 2014

[9] [9]

Beyond prediction error: 25 years of modeling the associations formed in the insect mushroom body.Learning & Memory, 31(5):a053824, 2024

Barbara Webb. Beyond prediction error: 25 years of modeling the associations formed in the insect mushroom body.Learning & Memory, 31(5):a053824, 2024

work page 2024

[10] [10]

Representations of novelty and familiarity in a mushroom body compartment.Cell, 169(5):956–969, 2017

Daisuke Hattori, Yoshinori Aso, Kurtis J Swartz, Gerald M Rubin, L F Abbott, and Richard Axel. Representations of novelty and familiarity in a mushroom body compartment.Cell, 169(5):956–969, 2017

work page 2017

[11] [11]

An anatomically constrained model for path integration in the bee brain.Current Biology, 27(20):3069–3085, 2017

Thomas Stone, Barbara Webb, Andrea Adden, Nicolai Ben Weddig, Anna Honkanen, Rachel Templin, William Wcislo, Luca Scimeca, Eric Warrant, and Stanley Heinze. An anatomically constrained model for path integration in the bee brain.Current Biology, 27(20):3069–3085, 2017

work page 2017

[12] [12]

The internal maps of insects.Journal of Experimental Biology, 222(Suppl_1):jeb188094, 2019

Barbara Webb. The internal maps of insects.Journal of Experimental Biology, 222(Suppl_1):jeb188094, 2019

work page 2019

[13] [13]

A connectome of theDrosophilacentral complex reveals network motifs suitable for flexible navigation and context-dependent action selection.eLife, 10, 2021

Brad K Hulse, Hannah Haberkern, Romain Franconville, Daniel Turner-Evans, Shin-ya Takemura, Tanya Wolff, Marcella Noorman, Marisa Dreher, Chuntao Dan, Ruchi Parekh, et al. A connectome of theDrosophilacentral complex reveals network motifs suitable for flexible navigation and context-dependent action selection.eLife, 10, 2021

work page 2021

[14] [14]

The central complex as a potential substrate for vector based navigation.Frontiers in Psychology, 10:690, 2019

Florent Le Moël, Thomas Stone, Mathieu Lihoreau, Antoine Wystrach, and Barbara Webb. The central complex as a potential substrate for vector based navigation.Frontiers in Psychology, 10:690, 2019

work page 2019

[15] [15]

A neurocomputational model of goal-directed navigation in insect-inspired artificial agents.Frontiers in Neurorobotics, 11:20, 2017

Dennis Goldschmidt, Poramate Manoonpong, and Sakyasingha Dasgupta. A neurocomputational model of goal-directed navigation in insect-inspired artificial agents.Frontiers in Neurorobotics, 11:20, 2017

work page 2017

[16] [16]

A decentralised neural model explaining optimal integration of navigational strategies in insects.eLife, 9:e54026, 2020

Xuelong Sun, Shigang Yue, and Michael Mangan. A decentralised neural model explaining optimal integration of navigational strategies in insects.eLife, 9:e54026, 2020

work page 2020

[17] [17]

Emergent spatial goals in an integrative model of the insect central complex.PLOS Computational Biology, 19(12):e1011480, 2023

Roman Goulard, Stanley Heinze, and Barbara Webb. Emergent spatial goals in an integrative model of the insect central complex.PLOS Computational Biology, 19(12):e1011480, 2023

work page 2023

[18] [18]

Neurons from pre-motor areas to the mushroom bodies can orchestrate latent visual learning in navigating insects.bioRxiv, 2023

Antoine Wystrach. Neurons from pre-motor areas to the mushroom bodies can orchestrate latent visual learning in navigating insects.bioRxiv, 2023

work page 2023

[19] [19]

Using the neural circuit of the insect central complex for path integration on a micro aerial vehicle

Jan Stankiewicz and Barbara Webb. Using the neural circuit of the insect central complex for path integration on a micro aerial vehicle. InConference on Biomimetic and Biohybrid Systems, pages 325–337. Springer, 2020

work page 2020

[20] [20]

A model of ant route navigation driven by scene familiarity.PLOS Computational Biology, 8(1):e1002336, 2012

Bart Baddeley, Paul Graham, Philip Husbands, and Andrew Philippides. A model of ant route navigation driven by scene familiarity.PLOS Computational Biology, 8(1):e1002336, 2012

work page 2012

[21] [21]

An insect-inspired model facilitating autonomous navigation by incorporating goal approaching and collision avoidance.Neural Networks, 165:106–118, 2023

Xuelong Sun, Qinbing Fu, Jigen Peng, and Shigang Yue. An insect-inspired model facilitating autonomous navigation by incorporating goal approaching and collision avoidance.Neural Networks, 165:106–118, 2023

work page 2023

[22] [22]

How might ants use panoramic views for route navigation?Journal of Experimental Biology, 214(3):445–451, 2011

Andrew Philippides, Bart Baddeley, Ken Cheng, and Paul Graham. How might ants use panoramic views for route navigation?Journal of Experimental Biology, 214(3):445–451, 2011

work page 2011

[23] [23]

The visual compass: Performance and limitations of an appearance-based method.Journal of Field Robotics, 23(10):913–941, 2006

Frédéric Labrosse. The visual compass: Performance and limitations of an appearance-based method.Journal of Field Robotics, 23(10):913–941, 2006

work page 2006

[24] [24]

Using an insect mush- room body circuit to encode route memory in complex natural environments.PLOS Computational Biology, 12(2):e1004683, 2016

Paul Ardin, Fei Peng, Michael Mangan, Konstantinos Lagogiannis, and Barbara Webb. Using an insect mush- room body circuit to encode route memory in complex natural environments.PLOS Computational Biology, 12(2):e1004683, 2016

work page 2016

[25] [25]

Route following without scanning

Aleksandar Kodzhabashev and Michael Mangan. Route following without scanning. In Stuart P. Wilson, Paul F.M.J. Verschure, Anna Mura, and Tony J. Prescott, editors,Biomimetic and Biohybrid Systems., pages 199–210, Cham, 2015. Springer International Publishing

work page 2015

[26] [26]

Familiarity-taxis: a bilateral approach to view-based snapshot navigation.Adaptive Behavior, 32(5):407–420, October 2024

Fabian Steinbeck, Efstathios Kagioulis, Alex Dewar, Andrew Philippides, Thomas Nowotny, and Paul Graham. Familiarity-taxis: a bilateral approach to view-based snapshot navigation.Adaptive Behavior, 32(5):407–420, October 2024. 18 APREPRINT- JANUARY26, 2026

work page 2024

[27] [27]

Insect-inspired embodied visual route following

Yihe Lu, Jiahao Cen, Rana Maroun Alkhoury, and Barbara Webb. Insect-inspired embodied visual route following. Journal of Bionic Engineering, 2025

work page 2025

[28] [28]

Insect visual homing strategies in a robot with analog processing.Biological Cybernetics, 83(3):231– 243, 2000

Ralf Möller. Insect visual homing strategies in a robot with analog processing.Biological Cybernetics, 83(3):231– 243, 2000

work page 2000

[29] [29]

Learning and time-dependent cue choice in the desert ant,Melophorus bagoti

Cody A Freas and Ken Cheng. Learning and time-dependent cue choice in the desert ant,Melophorus bagoti. Ethology, 123(8):503–515, 2017

work page 2017

[30] [30]

Terrestrial cue learning and retention during the outbound and inbound foraging trip in the desert ant, cataglyphis velox.Journal of Comparative Physiology A, 205(2):177–189, 2019

Cody A Freas and Marcia L Spetch. Terrestrial cue learning and retention during the outbound and inbound foraging trip in the desert ant, cataglyphis velox.Journal of Comparative Physiology A, 205(2):177–189, 2019

work page 2019

[31] [31]

Latent learning without map-like representation of space in navigating ants.bioRxiv, pages 2024–08, 2024

Leo Clement, Sebastian Schwarz, and Antoine Wystrach. Latent learning without map-like representation of space in navigating ants.bioRxiv, pages 2024–08, 2024

work page 2024

[32] [32]

iGibson 2.0: object-centric simulation for robot learning of everyday household tasks

Chengshu Li, Fei Xia, Roberto Martín-Martín, Michael Lingelbach, Sanjana Srivastava, Bokui Shen, Kent Elliott Vainio, Cem Gokmen, Gokul Dharan, Tanish Jain, Andrey Kurenkov, Karen Liu, Hyowon Gweon, Jiajun Wu, Li Fei-Fei, and Silvio Savarese. iGibson 2.0: object-centric simulation for robot learning of everyday household tasks. InProceedings of the 5th Co...

work page 2022

[33] [33]

Theory of edge detection.Proceedings of the Royal Society of London

David Marr and Ellen Hildreth. Theory of edge detection.Proceedings of the Royal Society of London. Series B. Biological Sciences, 207(1167):187–217, 1980

work page 1980

[34] [34]

The properties of the visual system in the Australian desert antMelophorus bagoti.Arthropod Structure & Development, 40(2):128–134, 2011

Sebastian Schwarz, Ajay Narendra, and Jochen Zeil. The properties of the visual system in the Australian desert antMelophorus bagoti.Arthropod Structure & Development, 40(2):128–134, 2011

work page 2011

[35] [35]

Predicting visual function by interpreting a neuronal wiring diagram.Nature, 634(8032):113– 123, 2024

H Sebastian Seung. Predicting visual function by interpreting a neuronal wiring diagram.Nature, 634(8032):113– 123, 2024

work page 2024

[36] [36]

Antoine Wystrach, Alex Dewar, Andrew Philippides, and Paul Graham. How do field of view and resolution affect the information content of panoramic scenes for visual navigation? A computational investigation.Journal of Comparative Physiology A, 202:87–95, 2016

work page 2016

[37] [37]

Route-centric ant-inspired memories enable panoramic route-following in a car-like robot.Nature Communications, 16(1):8328, 2025

Gabriel G Gattaux, Antoine Wystrach, Julien R Serres, and Franck Ruffier. Route-centric ant-inspired memories enable panoramic route-following in a car-like robot.Nature Communications, 16(1):8328, 2025

work page 2025

[38] [38]

Isotropic-sequence-order learning in a closed-loop behavioural system

Bernd Porr and Florentin Wörgötter. Isotropic-sequence-order learning in a closed-loop behavioural system. Philosophical Transactions of the Royal Society of London. Series A: Mathematical, Physical and Engineering Sciences, 361(1811):2225–2244, 2003

work page 2003

[39] [39]

Integration of parallel opposing memories underlies memory extinction.Cell, 175(3):709–722, 2018

Johannes Felsenberg, Pedro F Jacob, Thomas Walker, Oliver Barnstedt, Amelia J Edmondson-Stait, Markus W Pleijzier, Nils Otto, Philipp Schlegel, Nadiya Sharifi, Emmanuel Perisse, et al. Integration of parallel opposing memories underlies memory extinction.Cell, 175(3):709–722, 2018

work page 2018

[40] [40]

A neural data structure for novelty detection.Proceedings of the National Academy of Sciences, 115(51):13093–13098, 2018

Sanjoy Dasgupta, Timothy C Sheehan, Charles F Stevens, and Saket Navlakha. A neural data structure for novelty detection.Proceedings of the National Academy of Sciences, 115(51):13093–13098, 2018

work page 2018

[41] [41]

Traces of drosophila memory.Neuron, 70(1):8–19, 2011

Ronald L Davis. Traces of drosophila memory.Neuron, 70(1):8–19, 2011

work page 2011

[42] [42]

An incentive circuit for memory dynamics in the mushroom body ofDrosophila melanogaster.eLife, 11:e75611, 2022

Evripidis Gkanias, Li Yan McCurdy, Michael N Nitabach, and Barbara Webb. An incentive circuit for memory dynamics in the mushroom body ofDrosophila melanogaster.eLife, 11:e75611, 2022

work page 2022

[43] [43]

iGibson 1.0: a simulation environment for interactive tasks in large realistic scenes

Bokui Shen, Fei Xia, Chengshu Li, Roberto Martín-Martín, Linxi Fan, Guanzhi Wang, Claudia Pérez-D’Arpino, Shyamal Buch, Sanjana Srivastava, Lyne Tchapmi, et al. iGibson 1.0: a simulation environment for interactive tasks in large realistic scenes. In2021 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS), pages 7520–7527. IEEE, 2021

work page 2021

[44] [44]

Zamir, Zhi-Yang He, Alexander Sax, Jitendra Malik, and Silvio Savarese

Fei Xia, Amir R. Zamir, Zhi-Yang He, Alexander Sax, Jitendra Malik, and Silvio Savarese. Gibson Env: real-world perception for embodied agents. InComputer Vision and Pattern Recognition (CVPR), 2018 IEEE Conference on. IEEE, 2018

work page 2018

[45] [45]

Benchmarking Classic and Learned Navigation in Complex 3D Environments

Dmytro Mishkin, Alexey Dosovitskiy, and Vladlen Koltun. Benchmarking classic and learned navigation in complex 3D environments.arXiv preprint arXiv:1901.10915, 2019

work page internal anchor Pith review Pith/arXiv arXiv 1901

[46] [46]

The surprising effectiveness of visual odometry techniques for embodied pointgoal navigation

Xiaoming Zhao, Harsh Agrawal, Dhruv Batra, and Alexander Schwing. The surprising effectiveness of visual odometry techniques for embodied pointgoal navigation. In2021 IEEE/CVF International Conference on Computer Vision (ICCV), pages 16107–16116, 2021

work page 2021

[47] [47]

Is mapping necessary for realistic pointgoal navigation? In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17211–17220, 2022

Ruslan Partsey, Erik Wijmans, Naoki Yokoyama, Oles Dobosevych, Dhruv Batra, and Oleksandr Maksymets. Is mapping necessary for realistic pointgoal navigation? In2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), pages 17211–17220, 2022. 19 APREPRINT- JANUARY26, 2026

work page 2022

[48] [48]

Sim2real predictivity: does evaluation in simulation predict real-world performance?IEEE Robotics and Automation Letters, 5(4):6670–6677, 2020

Abhishek Kadian, Joanne Truong, Aaron Gokaslan, Alexander Clegg, Erik Wijmans, Stefan Lee, Manolis Savva, Sonia Chernova, and Dhruv Batra. Sim2real predictivity: does evaluation in simulation predict real-world performance?IEEE Robotics and Automation Letters, 5(4):6670–6677, 2020

work page 2020

[49] [49]

MINOS: Multimodal Indoor Simulator for Navigation in Complex Environments

Manolis Savva, Angel X. Chang, Alexey Dosovitskiy, Thomas Funkhouser, and Vladlen Koltun. MINOS: multimodal indoor simulator for navigation in complex environments.arXiv:1712.03931, 2017

work page internal anchor Pith review Pith/arXiv arXiv 2017

[50] [50]

Cognitive architecture of a mini-brain: the honeybee.Trends in Cognitive Sciences, 5(2):62–71, 2001

Randolf Menzel and Martin Giurfa. Cognitive architecture of a mini-brain: the honeybee.Trends in Cognitive Sciences, 5(2):62–71, 2001

work page 2001

[51] [51]

Allometric analysis of brain cell num- ber in Hymenoptera suggests ant brains diverge from general trends.Proceedings of the Royal Society B, 288(1947):20210199, 2021

R Keating Godfrey, Mira Swartzlander, and Wulfila Gronenberg. Allometric analysis of brain cell num- ber in Hymenoptera suggests ant brains diverge from general trends.Proceedings of the Royal Society B, 288(1947):20210199, 2021

work page 1947

[52] [52]

A lateralised design for the interaction of visual memories and heading representations in navigating ants.bioRxiv, 2020

Antoine Wystrach, Florent Le Moël, Leo Clement, and Sebastian Schwarz. A lateralised design for the interaction of visual memories and heading representations in navigating ants.bioRxiv, 2020

work page 2020

[53] [53]

Obstacle avoidance in a crowd with a plastic recurrent mushroom body model.IEEE Access, 2025

Liying Tao, Zonglin Yang, Gaoming Li, and Delong Shang. Obstacle avoidance in a crowd with a plastic recurrent mushroom body model.IEEE Access, 2025

work page 2025

[54] [54]

A comparative study of bug algorithms for robot navigation.Robotics and Autonomous Systems, 121:103261, 2019

Kimberly N McGuire, Guido C H E de Croon, and Karl Tuyls. A comparative study of bug algorithms for robot navigation.Robotics and Autonomous Systems, 121:103261, 2019

work page 2019

[55] [55]

Lumelsky and A

V . Lumelsky and A. Stepanov. Dynamic path planning for a mobile automaton with limited information on the environment.IEEE Transactions on Automatic Control, 31(11):1058–1063, 1986

work page 1986

[56] [56]

Learning with reinforcement prediction errors in a model of theDrosophilamushroom body.Nature Communications, 12(1):2569, 2021

James EM Bennett, Andrew Philippides, and Thomas Nowotny. Learning with reinforcement prediction errors in a model of theDrosophilamushroom body.Nature Communications, 12(1):2569, 2021

work page 2021

[57] [57]

Prediction error drives associative learning and conditioned behavior in a spiking model of Drosophila larva

Anna-Maria Jürgensen, Panagiotis Sakagiannis, Michael Schleyer, Bertram Gerber, and Martin Paul Nawrot. Prediction error drives associative learning and conditioned behavior in a spiking model of Drosophila larva. iScience, 27(1), 2024

work page 2024

[58] [58]

Stephan Lochner, Daniel Honerkamp, Abhinav Valada, and Andrew D Straw. Reinforcement learning as a robotics- inspired framework for insect navigation: from spatial representations to neural implementation.Frontiers in Computational Neuroscience, 18:1460006, 2024

work page 2024

[59] [59]

Sleep deprivation results in diverse patterns of synaptic scaling across theDrosophilamushroom bodies.Current Biology, 31(15):3248–3261, 2021

Jacqueline T Weiss and Jeffrey M Donlea. Sleep deprivation results in diverse patterns of synaptic scaling across theDrosophilamushroom bodies.Current Biology, 31(15):3248–3261, 2021

work page 2021

[60] [60]

Catastrophic forgetting in connectionist networks.Trends in Cognitive Sciences, 3(4):128–135, 1999

Robert M French. Catastrophic forgetting in connectionist networks.Trends in Cognitive Sciences, 3(4):128–135, 1999

work page 1999

[61] [61]

Processing of expected and unexpected events during conditioning and attention: a psy- chophysiological theory.Psychological Review, 89(5):529, 1982

Stephen Grossberg. Processing of expected and unexpected events during conditioning and attention: a psy- chophysiological theory.Psychological Review, 89(5):529, 1982

work page 1982

[62] [62]

Reducing catastrophic forgetting with associative learning: a lesson from fruit flies.Neural Computation, 35(11):1797–1819, 2023

Yang Shen, Sanjoy Dasgupta, and Saket Navlakha. Reducing catastrophic forgetting with associative learning: a lesson from fruit flies.Neural Computation, 35(11):1797–1819, 2023

work page 2023

[63] [63]

Large scale homing in honeybees.PLOS One, 6(5):e19669, 2011

Mario Pahl, Hong Zhu, Jürgen Tautz, and Shaowu Zhang. Large scale homing in honeybees.PLOS One, 6(5):e19669, 2011

work page 2011

[64] [64]

Ants might use different view-matching strategies on and off the route.Journal of Experimental Biology, 215(1):44–55, 2012

Antoine Wystrach, Guy Beugnon, and Ken Cheng. Ants might use different view-matching strategies on and off the route.Journal of Experimental Biology, 215(1):44–55, 2012

work page 2012

[65] [65]

Mushroom body output neurons encode valence and guide memory-based action selection in drosophila.eLife, 3:e04580, 2014

Yoshinori Aso, Divya Sitaraman, Toshiharu Ichinose, Karla R Kaun, Katrin V ogt, Ghislain Belliart-Guérin, Pierre-Yves Plaçais, Alice A Robie, Nobuhiro Yamagata, Christopher Schnaitmann, et al. Mushroom body output neurons encode valence and guide memory-based action selection in drosophila.eLife, 3:e04580, 2014

work page 2014

[66] [66]

Opponent processes in visual memories: A model of attraction and repulsion in navigating insects’ mushroom bodies.PLOS Computational Biology, 16(2):e1007631, 2020

Florent Le Möel and Antoine Wystrach. Opponent processes in visual memories: A model of attraction and repulsion in navigating insects’ mushroom bodies.PLOS Computational Biology, 16(2):e1007631, 2020

work page 2020

[67] [67]

A comparative study of reinforcement learning and insect-inspired visual navigation methods

Xiaoting Zhong, Xuelong Sun, and Haiyang Li. A comparative study of reinforcement learning and insect-inspired visual navigation methods. InConference on Biomimetic and Biohybrid Systems, pages 163–175. Springer, 2025. 20 APREPRINT- JANUARY26, 2026 5 Appendix 5.1 Model Details See Table 3. Table 3: More details of the recent point-goal navigation models i...

work page 2025

[68] [68]

SunCG, Matterport3D ORB-SLAM2, D* Lite — CNN

work page

[69] [69]

Gibson 4+, Matterport3D PPO/inherited from [45] SR++, geodesic-, frame- CNN

work page

[70] [70]

Gibson 2+, Matterport3D DDPPO SPL++, geodesic-, frame- SE-ResNeXt50

work page

[71] [71]

Gibson 4+ inherited from [3] inherited from [3] ResNet18

work page

[72] [72]

Gibson 0+ inherited from [3] inherited from [3] ResNet50

work page

[73] [73]

HM3D inherited from [3] inherited from [3] ResNet50 5.2 Supplementary Figures See Fig. 10-12. Figure 10: Zoom-in plots of the area enclosed in the rectangle with dashed boundaries in Fig. 6. The black circles here denote where collisions were detected. 21 APREPRINT- JANUARY26, 2026 Figure 11: The effect of motor noise on performance, across trials within ...

work page 2026