The spiking output of the mouse olfactory bulb encodes large-scale temporal features of natural odor environments

In natural odor environments, odor travels in plumes. Odor concentration dynamics change in characteristic ways across the width and length of a plume. Thus, spatiotemporal dynamics of plumes have informative features for animals navigating to an odor source. Population activity in the olfactory bulb (OB) has been shown to follow odor concentration across plumes to a moderate degree (Lewis et al., 2021). However, it is unknown whether the ability to follow plume dynamics is driven by individual cells or whether it emerges at the population level. Previous research has explored the responses of individual OB cells to isolated features of plumes, but it is difficult to adequately sample the full feature space of plumes as it is still undetermined which features navigating mice employ during olfactory guided search. Here we released odor from an upwind odor source and simultaneously recorded both odor concentration dynamics and cellular response dynamics in awake, head-fixed mice. We found that longer timescale features of odor concentration dynamics were encoded at both the cellular and population level. At the cellular level, responses were elicited at the beginning of the plume for each trial, signaling plume onset. Plumes with high odor concentration elicited responses at the end of the plume, signaling plume offset. Although cellular level tracking of plume dynamics was observed to be weak, we found that at the population level, OB activity distinguished whiffs and blanks (accurately detected odor presence versus absence) throughout the duration of a plume. Even ~20 OB cells were enough to accurately discern odor presence throughout a plume. Our findings indicate that the full range of odor concentration dynamics and high frequency fluctuations are not encoded by OB spiking activity. Instead, relatively lower-frequency temporal features of plumes, such as plume onset, plume offset, whiffs, and blanks, are represented in the OB.

A vast space of compact strategies for effective decisions

Inference-based decision-making, which underlies a broad range of behavioral tasks, is typically studied using a small number of handcrafted models. We instead enumerate a complete ensemble of strategies that could be used to effectively, but not necessarily optimally, solve a dynamic foraging task. Each strategy is expressed as a behavioral "program" that uses a limited number of internal states to specify actions conditioned on past observations. We show that the ensemble of strategies is enormous-comprising a quarter million programs with up to five internal states-but can nevertheless be understood in terms of algorithmic "mutations" that alter the structure of individual programs. We devise embedding algorithms that reveal how mutations away from a Bayesian-like strategy can diversify behavior while preserving performance, and we construct a compositional description to link low-dimensional changes in algorithmic structure with high-dimensional changes in behavior. Together, this work provides an alternative approach for understanding individual variability in behavior across animals and tasks.

Adaptive Space-Aware Infotaxis II as a Strategy for Odor Source Localization

Mobile robot olfaction of toxic and hazardous odor sources is of great significance in anti-terrorism, disaster prevention, and control scenarios. Aiming at the problems of low search efficiency and easily falling into a local optimum of the current odor source localization strategies, the paper proposes the adaptive space-aware Infotaxis II algorithm. To improve the tracking efficiency of robots, a new reward function is designed by considering the space information and emphasizing the exploration behavior of robots. Considering the enhancement in exploratory behavior, an adaptive navigation-updated mechanism is proposed to adjust the movement range of robots in real time through information entropy to avoid an excessive exploration behavior during the search process, which may lead the robot to fall into a local optimum. Subsequently, an improved adaptive cosine salp swarm algorithm is applied to confirm the optimal information adaptive parameter. Comparative simulation experiments between ASAInfotaxis II and the classical search strategies are carried out in 2D and 3D scenarios regarding the search efficiency and search behavior, which show that ASAInfotaxis II is competent to improve the search efficiency to a larger extent and achieves a better balance between exploration and exploitation behaviors.

Optimal policies for Bayesian olfactory search in turbulent flows

In many practical scenarios, a flying insect must search for the source of an emitted cue which is advected by the atmospheric wind. On the macroscopic scales of interest, turbulence tends to mix the cue into patches of relatively high concentration over a background of very low concentration, so that the insect will detect the cue only intermittently and cannot rely on chemotactic strategies which simply climb the concentration gradient. In this work we cast this search problem in the language of a partially observable Markov decision process and use the Perseus algorithm to compute strategies that are near-optimal with respect to the arrival time. We test the computed strategies on a large two-dimensional grid, present the resulting trajectories and arrival time statistics, and compare these to the corresponding results for several heuristic strategies, including (space-aware) infotaxis, Thompson sampling, and QMDP. We find that the near-optimal policy found by our implementation of Perseus outperforms all heuristics we test by several measures. We use the near-optimal policy to study how the search difficulty depends on the starting location. We also discuss the choice of initial belief and the robustness of the policies to changes in the environment. Finally, we present a detailed and pedagogical discussion about the implementation of the Perseus algorithm, including the benefits-and pitfalls-of employing a reward-shaping function.

Deep reinforcement learning for the olfactory search POMDP: a quantitative benchmark

The olfactory search POMDP (partially observable Markov decision process) is a sequential decision-making problem designed to mimic the task faced by insects searching for a source of odor in turbulence, and its solutions have applications to sniffer robots. As exact solutions are out of reach, the challenge consists in finding the best possible approximate solutions while keeping the computational cost reasonable. We provide a quantitative benchmarking of a solver based on deep reinforcement learning against traditional POMDP approximate solvers. We show that deep reinforcement learning is a competitive alternative to standard methods, in particular to generate lightweight policies suitable for robots.