The oneirogen hypothesis: modeling the hallucinatory effects of classical psychedelics in terms of replay-dependent plasticity mechanisms

  1. Colin Bredenberg
  2. Fabrice Normandin
  3. Blake Richards
  4. Guillaume Lajoie

  • Curated by eLife

    eLife logo

    eLife Assessment

    This paper provides a useful new theory of the hallucinatory effects of psychedelics. The authors present convincing evidence that a computational model trained with the Wake-Sleep algorithm can reproduce some features of hallucinations by varying the strength of top-down connections in the model, but discussion of the model's relationships to the psychedelics and sleep literatures is incomplete. The work will be of interest to researchers studying hallucinations or offline activity and learning more broadly.

Reviewed by

Read the full article See related articles

Listed in

Log in to save this article

Abstract

Abstract

Classical psychedelics induce complex visual hallucinations in humans, generating percepts that are co-herent at a low level, but which have surreal, dream-like qualities at a high level. While there are many hypotheses as to how classical psychedelics could induce these effects, there are no concrete mechanistic models that capture the variety of observed effects in humans, while remaining consistent with the known pharmacological effects of classical psychedelics on neural circuits. In this work, we propose the “oneirogen hypothesis”, which posits that the perceptual effects of classical psychedelics are a result of their pharmacological actions inducing neural activity states that truly are more similar to dream-like states. We simulate classical psychedelics’ effects via manipulating neural network models trained on perceptual tasks with the Wake-Sleep algorithm. This established machine learning algorithm leverages two activity phases, a perceptual phase (wake) where sensory inputs are encoded, and a generative phase (dream) where the network internally generates activity consistent with stimulus-evoked responses. We simulate the action of psychedelics by partially shifting the model to the ‘Sleep’ state, which entails a greater influence of top-down connections, in line with the impact of psychedelics on apical dendrites. The effects resulting from this manipulation capture a number of experimentally observed phenomena including the emergence of hallucinations, increases in stimulus-conditioned variability, and large increases in synaptic plasticity. We further provide a number of testable predictions which could be used to validate or invalidate our oneirogen hypothesis.

Article activity feed

  1. eLife

    eLife Assessment

    This paper provides a useful new theory of the hallucinatory effects of psychedelics. The authors present convincing evidence that a computational model trained with the Wake-Sleep algorithm can reproduce some features of hallucinations by varying the strength of top-down connections in the model, but discussion of the model's relationships to the psychedelics and sleep literatures is incomplete. The work will be of interest to researchers studying hallucinations or offline activity and learning more broadly.

  2. eLife

    Reviewer #1 (Public review):

    Bredenberg et al. aim to model some of the visual and neural effects of psychedelics via the Wake-Sleep algorithm. This is an interesting study with findings that go against certain mainstream ideas in psychedelic neuroscience (that I largely agree with). I cannot speak to the math in this manuscript, but it seems like quite a conceptual leap to set a parameter of the model in between wake and sleep and state that this is a proxy to acute psychedelic effects (point #20). My other concerns below are related to the review of the psychedelic literature:

    (1) Page 1, Introduction, "...they are agonists for the 5-HT2a serotonin receptor commonly expressed on the apical dendrites of cortical pyramidal neurons..." It is a bit redundant to say "5-HT2A serotonin receptor," as serotonin is already captured by its abbreviation (i.e., 5-HT).

    While psychedelic research has focused on 5-HT2A expression on cortical pyramidal cells, note that the 5-HT2A receptor is also expressed on interneurons in the medial temporal lobe (entorhinal cortex, hippocampus, and amygdala) with some estimates being >50% of these neurons (https://doi.org/10.1016/j.brainresbull.2011.11.006, https://doi.org/10.1007/s00221-013-3512-6, https://doi.org/10.7554/eLife.66960, https://doi.org/10.1016/j.mcn.2008.07.005, https://doi.org/10.1038/npp.2008.71, https://doi.org/10.1038/s41386-023-01744-8, https://doi.org/10.1016/j.brainres.2004.03.016, https://doi.org/10.1016/S0022-3565(24)37472-5, https://doi.org/10.1002/hipo.22611, https://doi.org/10.1016/j.neuron.2024.08.016). However, with ~1:4 ratio of inhibitory to excitatory neurons in the brain (https://doi.org/10.1101/2024.09.24.614724), this can make it seem as if 5-HT2A expression is negligible in the MTL. I think it might be important to mention these receptors, as this manuscript discusses replay.

    I see now that Figure 1 mentions that PV cells also express 5-HT2A receptors. This should probably be mentioned earlier.

    (2) Page 1, Introduction, "They have further been used for millennia as medicine and in religious rituals..." This might be a romanticization of psychedelics and indigenous groups, as anthropological evidence suggests that intentional psychedelic use might actually be more recent (see work by Manvir Singh and Andy Letcher).

    (3) When discussing oneirogens, it could be worth differentiating psychedelics from kappa opioid agonists such as ibogaine and salvinorin A, another class of hallucinogens that some refer to as "oneirogens" (similar to how "psychedelic" is the colloquial term for 5-HT2A agonists). Note that studies have found the effects of Salvia divinorum (which contains salvinorin A) to be described more similarly to dreams than psychedelics (https://doi.org/10.1007/s00213-011-2470-6). This makes me wonder why the present study is more applicable to 5-HT2A psychedelics than other kappa opioid agonists or other classes of hallucinogens (e.g., NMDA antagonists, muscarinic antagonists, GABAA agonists).

    (4) Page 2, Introduction, "Replay sequences have been shown to be important for learning during sleep [14, 15, 16, 17, 18]: we propose that mechanisms supporting replay-dependent learning during sleep are key to explaining the increases in plasticity caused by psychedelic drug administration." I'm not sure I follow the logic of this point. Dreams happen during REM sleep, whereas replay is most prominent during non-REM sleep. Moreover, while it's not clear what psychedelics do to hippocampal function, most evidence would suggest they impair it. As mentioned, most 5-HT2A receptors in the hippocampus seem to be on inhibitory neurons, and human and animal work finds that psychedelics impair hippocampal-dependent memory encoding (https://doi.org/10.1037/rev0000455, https://doi.org/10.1037/rev0000455, https://doi.org/10.3389/fnbeh.2014.00180, https://doi.org/10.1002/hipo.22712). One study even found that psilocin impairs hippocampal-dependent memory retrieval (https://doi.org/10.3389/fnbeh.2014.00180). Note that this is all in reference to the acute effects (psychedelics may post-acutely enhance hippocampal-dependent memory, https://doi.org/10.1007/s40265-024-02106-4).

    (5) Page 2, Introduction, "In total, our model of the functional effect of psychedelics on pyramidal neurons could provide a explanation for the perceptual psychedelic experience in terms of learning mechanisms for consolidation during sleep..." In contrast to my previous point, I think this could be possible. Three datasets have found that psychedelics may enhance cortical-dependent memory encoding (i.e., familiarity; https://doi.org/10.1037/rev0000455, https://doi.org/10.1037/rev0000455), and two studies found that post-encoding administration of psychedelics retroactively enhanced memory that may be less hippocampal-dependent/more cortical-dependent (https://doi.org/10.1016/j.neuropharm.2012.06.007, https://doi.org/10.1016/j.euroneuro.2022.01.114). Moreover, and as mentioned below, 5 studies have found decoupling between the hippocampus and the cortex (https://doi.org/10.3389/fnhum.2014.00020, https://doi.org/10.1002/hbm.22833, https://doi.org/10.1016/j.celrep.2021.109714, https://doi.org/10.1162/netn_a_00349, https://doi.org/10.1038/s41586-024-07624-5), something potentially also observed during REM sleep that is thought to support consolidation (https://doi.org/10.1073/pnas.2123432119). These findings should probably be discussed.

    (6) Page 2, Introduction, "In this work, we show that within a neural network trained via Wake-Sleep, it is possible to model the action of classical psychedelics (i.e. 5-HT2a receptor agonism)..." Note that 5-HT2A agonism alone is not sufficient to explain the effects of psychedelics, given that there are 5-HT2A agonists that are non-hallucinogenic (e.g., lisuride).

    (7) Page 2, Introduction, "...by shifting the balance during the wake state from the bottom-up pathways to the top-down pathways, thereby making the 'wake' network states more 'dream-like'." I could have included this in the previous point, but I felt that this idea deserved its own point. There has been a rather dogmatic assertion that psychedelics diminish top-down processing and/or enhance bottom-up processing, and I appreciate that the authors have not accepted this as fact. However, because this is an unfortunately prominent idea, I think it ought to be fleshed out more by first mentioning that it's one of the tenets of REBUS. REBUS has become a popular model of psychedelic drug action, but it's largely unfalsifiable (it's based on two unfalsifiable models, predictive processing and integrated information theory), so the findings from this study could tighten it up a bit. Second, there have now been a handful of studies that have attempted to study directionality in information flow under psychedelics, and the findings are rather mixed including increased bottom-up/decreased top-down effects (https://doi.org/10.7554/eLife.59784, https://doi.org/10.1073/pnas.1815129116; note that the latter "bottom-up" effect involves subcortical-cortical connections in which it's less clear what's actually "higher-/lower-level"), increased top-down/decreased bottom-up effects (https://doi.org/10.1038/s41380-024-02632-3, https://doi.org/10.1016/j.euroneuro.2016.03.018), or both (https://doi.org/10.1016/j.neuroimage.2019.116462, https://doi.org/10.1016/j.neuropharm.2017.10.039), though most of these studies are aggregating across largely inhomogeneous states (i.e., resting-state). Lastly, and somewhat problematically, facilitated top-down processing is also an idea proposed in psychosis that's based partially on findings with acute ketamine administration (note that all hallucinations to some degree might rely on top-down facilitation, as a hallucination involves a high-level concept that impinges on lower-level sensory areas; see work by Phil Corlett). While psychosis and the effects of ketamine have some similarities with psychedelics, there are certainly differences, and I think the goal of this manuscript is to uniquely describe 5-HT2A psychedelics (again, I'm left wondering why tweaking alpha in the Wake-Sleep algorithm is any more applicable to psychedelics than other hallucinogenic conditions).

    (8) Figure 2 equates alpha with a "psychedelic dose," but this is a bit misleading, as neither the algorithm nor an individual was administered a psychedelic. Alpha is instead a hypothetical proxy for a psychedelic dose. Moreover, if the model were recapitulating the effects of psychedelics, shouldn't these images look more psychedelic as alpha increases (e.g., they may look like images put through the DeepDream algorithm).

    (9) Page 11, Methods, "...and the gate α ensures that learning only occurs during sleep mode... The (1 − α) gate in this case ensures that plasticity only occurs during the Wake mode." Much of the math escapes me, so perhaps I'm misunderstanding these statements, but learning and plasticity certainly happen during both wake and sleep, making me wonder what is meant by these statements. Moreover, if plasticity is simply neural changes, couldn't plasticity be synonymous with neural learning? Perhaps plasticity and learning are meant to refer to different types of neural changes. It might be worth clarifying this, as a general problem in psychedelic research is that psychedelics are described as facilitating plasticity when brains are changing at every moment (hence not experiencing every moment as the same), and psychedelics don't impact all forms of plasticity equally. For example, psychedelics may not necessarily enhance neurogenesis or the addition of certain receptor types, and they impair certain forms of learning (i.e., episodic memory encoding). What is typically meant by plasticity enhancements induced by psychedelics (and where there's the most evidence) is dendritic plasticity (i.e., the growth of dendrites and spines). Whatever is meant by "plasticity" should be clarified in its first instance in this manuscript.

    (10) Page 12, Methods, "During training, neural network activity is either dominated entirely by bottom-up inputs (Wake, α = 0) or by top-down inputs (Sleep, α = 1)." Again, I could be misunderstanding the mathematical formulation, but top-down inputs operate during wake, and bottom-up inputs can operate during sleep (people can wake up or even incorporate noise from their environments into sleep.

    (11) Page 4, Results, "Thus, we can capture the core idea behind the oneirogen hypothesis using the Wake-Sleep algorithm, by postulating that the bottom-up basal synapses are predominantly driving neural activity during the Wake phase (when α is low)." However, several pieces of evidence (and the first circuit model of psychedelic drug action) suggest that psychedelics enhance functional connectivity and potentially even effective connectivity from the thalamus to the cortex (https://doi.org/10.1093/brain/awab406). Note that psychedelics may not equally impact all subcortical structures. REBUS proposes the opposite of the current study, that psychedelics facilitate bottom-up information flow, with one of the few explicit predictions being that psychedelics should facilitate information flow from the hippocampus to the default mode network. However, as mentioned earlier, 5 studies have found that psychedelics diminish functional connectivity between the hippocampus and cortex (including the DMN but also V1).

    (12) Page 4, Results, "...and have an excitatory effect that positively modulates glutamatergic transmission..." Note that this may not be brainwide. While psychedelics were found to increase glutamatergic transmission in the cortex, they were also found to decrease hippocampal glutamate (consistent with inhibition of the hippocampus, https://doi.org/10.1038/s41386-020-0718-8).

    (13) Page 5, "...which are similar to the 'breathing' and 'rippling' phenomena reported by psychedelic drug users at low doses..." Although it's sometimes unclear what is meant by "low doses," the breathing/rippling effect of psychedelics occurs at moderate and high doses as well.

    (14) I watched the videos, and it's hard for me to say there was some stark resemblance to psychedelic imagery. In contrast, for example, when the DeepDream algorithm came out, it did seem to capture something quite psychedelic.

    (15) Page 5, "This form of strongly correlated tuning has been observed in both cortex and the hippocampus." If this has been observed under non-psychedelic conditions, what does this tell us about this supposed model of psychedelics?

    (16) Page 6, with regards to neural variability, "...but whether this phenomenon [increased variability] is general across tasks and cortical areas remains to be seen." First, is variability here measured as variance? In fMRI datasets that have been used to support the Entropic Brain Hypothesis, note that variance tends to decrease, though certain measures of entropy increase (e.g., Figure 4A here https://doi.org/10.1073/pnas.1518377113 shows global variance decreases, and this reanalysis of those data https://doi.org/10.1002/hbm.23234 finds some entropy increases). Thus, variance and entropy should not be confused (in theory, one could cycle through several more brain states that are however, similar to each other, which would produce more entropy with decreased variance). Second, and perhaps more problematically for the EBH, is that the entropy effects of psychedelics completely disappear when one does a task, and unfortunately, the authors of these findings have misinterpreted them. What they'll say is that engaging in boring cognitive tasks or watching a video decreases entropy under psychedelics, but what you can see in Figure 1b of https://doi.org/10.1021/acschemneuro.3c00289 and Figure 4b of https://doi.org/10.1038/s41586-024-07624-5 is that entropy actually increases under sober conditions when you do a task. That is, it's a rather boring finding. Essentially, when resting in a scanner while sober, many may actually rest (including falling asleep, especially when subjects are asked to keep their eyes closed), and if you perform a task, brain activity should become more complex relative to doing nothing/falling asleep. When under a psychedelic, one can't fall asleep and thus, there's less change (though note that both of the above studies found numerical increases when performing tasks). Lastly, again I should note that the findings of the present study actually go against EBH/REBUS, given that the findings are increased top-down effects when EBH/REBUS predicts decreased top-down/increased bottom-up effects.

    (17) Page 6, "Because psychedelic drug administration increases influence of apical dendritic inputs on neural activity in our model, we found that silencing apical dendritic activity reduced across stimulus neural variability more as the psychedelic drug dose increases." I again want to point out that alpha is not the equivalent of a psychedelic dose here, but rather a parameter in the model that is being proposed as a proxy.

    (18) Page 8, "Experimentally, plasticity dynamics which could, theoretically, minimize such a prediction error have been observed in cortex [66, 67], and it has also been proposed that behavioral timescale plasticity in the hippocampus could subserve a similar function [68]. We found that plasticity rules of this kind induce strong correlations between inputs to the apical and basal dendritic compartments of pyramidal neurons, which have been observed in the hippocampus and cortex [55, 56]." Note that the plasticity effects of psychedelics are sometimes not observed in the hippocampus or are even observed as decreases (reviewed in https://doi.org/10.1038/s41386-022-01389-z).

    (19) Page 9, as is mentioned, REBUS proposes that there should be a decrease in top-down effects under psychedelics, which goes against what is found here, but as I describe above, the effects of psychedelics on various measures of directionality have been quite mixed.

    (20) Unless I'm misunderstanding something, it seems to be a bit of a jump to infer that simply changing alpha in your model is akin to psychedelic dosing. Perhaps if the model implemented biologically plausible 5-HT2A expression and/or its behavior were constrained by common features of a psychedelic experience (e.g., fractal-like visuals imposed onto perception, inability to fall asleep, etc.), I'd be more inclined to see the parallels between alpha and psychedelics dosing. However, it would still need to recapitulate unique effects of psychedelics (e.g., impairments in hippocampal-dependent memory with sparing/facilitation of cortical memory). At the moment, it seems like whatever the model is doing is applicable to any hallucinogenic drug or even psychosis.

  3. eLife

    Reviewer #2 (Public review):

    This work is a nice contribution to the literature in articulating a specific, testable theory of how psychedelics act to generate hallucinations and plasticity. The connection to replay, however - including in the title, abstract, and framing throughout the paper - is not well fleshed out.

    In particular, the paper's framing seems to conflate replay, dreams, and top-down processing, but these are not one and the same. Picard-Delano et al. TICS 2023 provides a useful review of the differences between replay and dreams. One key point is that most replay has been observed during NREM sleep, but our canonically bizarre / vivid dreams occur during REM. Top-down connections have also been proposed to be used for many processes aside from replay. The paper would benefit from much more precision and nuance on these points.

    I believe the paper is missing demonstrations or speculation about how plasticity under various doses of psychedelics relates to changes in performance, which would be an important link to the replay-dependent learning literature.

    Are there renderings available for 'ripple' effects of psychedelics that could be included, to allow readers to compare the model's hallucinations to humans'? Short of this, it would be useful to have a more detailed description of what rippling is. (For those readers without firsthand knowledge!) It is currently difficult to assess how close the match is.

  4. eLife

    Author response:

    We thank the reviewers for the valuable and constructive reviews. Thanks to these, we believe the article will be considerably improved. We have organized our response to address points that are relevant to both reviewers first, after which we address the unique concerns of each individual reviewer separately. We briefly paraphrase each concern and provide comments for clarification, outlining the precise changes that we will make to the text.

    Common Concerns (Reviewer 1 & Reviewer 2):

    Can you clarify how NREM and REM sleep relate to the oneirogen hypothesis?

    Within the submission draft we tried to stay agnostic as to whether mechanistically similar replay events occur during NREM or REM sleep; however, upon a more thorough literature review, we think that there is moderately greater evidence in favor of Wake-Sleep-type replay occurring during REM sleep which is related to classical psychedelic drug mechanisms of action.

    First, we should clarify that replay has been observed during both REM and NREM sleep, and dreams have been documented during both sleep stages, though the characteristics of dreams differ across stages, with NREM dreams being more closely tied to recent episodic experience and REM dreams being more bizarre/hallucinatory (see Stickgold et al., 2001 for a review). Replay during sleep has been studied most thoroughly during NREM sharp-wave ripple events, in which significant cortical-hippocampal coupling has been observed (Ji & Wilson, 2007). However, it is critical to note that the quantification methods used to identify replay events in the hippocampal literature usually focus on identifying what we term ‘episodic replay,’ which involves a near-identical recapitulation of neural trajectories that were recently experienced during waking experimental recordings (Tingley & Peyrach, 2020). In contrast, our model focuses on ‘generative replay,’ where one expects only a statistically similar reproduction of neural activity, without any particular bias towards recent or experimentally controlled experience. This latter form of replay may look closer to the ‘reactivation’ observed in cortex by many studies (e.g. Nguyen et al., 2024), where correlation structures of neural activity similar to those observed during stimulus-driven experience are recapitulated. Under experimental conditions in which an animal is experiencing highly stereotyped activity repeatedly, over extended periods of time, these two forms of replay may be difficult to dissociate.

    Interestingly, though NREM replay has been shown to couple hippocampal and cortical activity, a similar study in waking animals administered psychedelics found hippocampal replay without any obvious coupling to cortical activity (Domenico et al., 2021). This could be because the coupling was not strong enough to produce full trajectories in the cortex (psychedelic administration did not increase ‘alpha’ enough), and that a causal manipulation of apical/basal influence in the cortex may be necessary to observe the increased coupling. Alternatively, as Reviewer 1 noted, it may be that psychedelics induce a form of hippocampus-decoupled replay, as one would expect from the REM stage of a recently proposed complementary learning systems model (Singh et al., 2022).

    Evidence in favor of a similarity between the mechanism of action of classical psychedelics and the mechanism of action of memory consolidation/learning during REM sleep is actually quite strong. In particular, studies have shown that REM sleep increases the activity of soma-targeting parvalbumin (PV) interneurons and decreases the activity of apical dendrite-targeting somatostatin (SOM) interneurons (Niethard et al., 2021), that this shift in balance is controlled by higher-order thalamic nuclei, and that this shift in balance is critical for synaptic consolidation of both monocular deprivation effects in early visual cortex (Zhou et al., 2020) and for the consolidation of auditory fear conditioning in the dorsal prefrontal cortex (Aime et al., 2022). These last studies were not discussed in the present manuscript–we will add them, in addition to a more nuanced description of the evidence connecting our model to NREM and REM replay.

    Can you explain how synaptic plasticity induced by psychedelics within your model relates to learning at a behavioral level?

    While the Wake-Sleep algorithm is a useful model for unsupervised statistical learning, it is not a model of reward or fear-based conditioning, which likely occur via different mechanisms in the brain (e.g. dopamine-dependent reinforcement learning or serotonin-dependent emotional learning). The Wake-Sleep algorithm is a ‘normative plasticity algorithm,’ that connects synaptic plasticity to the formation of structured neural representations, but it is not the case that all synaptic plasticity induced by psychedelic administration within our model should induce beneficial learning effects. According to the Wake-Sleep algorithm, plasticity at apical synapses is enhanced during the Wake phase, and plasticity at basal synapses is enhanced during the Sleep phase; under the oneirogen hypothesis, hallucinatory conditions (increased ‘alpha’) cause an increase in plasticity at both apical and basal sites. Because neural activity is in a fundamentally aberrant state when ‘alpha’ is increased, there are no theoretical guarantees that plasticity will improve performance on any objective: psychedelic-induced plasticity within our model could perhaps better be thought of as ‘noise’ that may have a positive or negative effect depending on the context.

    In particular, such ‘noise’ may be beneficial for individuals or networks whose synapses have become locked in a suboptimal local minimum. The addition of large amounts of random plasticity could allow a system to extricate itself from such local minima over subsequent learning (or with careful selection of stimuli during psychedelic experience), similar to simulated annealing optimization approaches. If our model were fully validated, this view of psychedelic-induced plasticity as ‘noise’ could have relevance for efforts to alleviate the adverse effects of PTSD, early life trauma, or sensory deprivation; it may also provide a cautionary note against repeated use of psychedelic drugs within a short time frame, as the plasticity changes induced by psychedelic administration under our model are not guaranteed to be good or useful in-and-of themselves without subsequent re-learning and compensation.

    We should also note that we have deliberately avoided connecting the oneirogen hypothesis model to fear extinction experimental results that have been observed through recordings of the hippocampus or the amygdala (Bombardi & Giovanni, 2013; Jiang et al., 2009; Kelly et al., 2024; Tiwari et al., 2024). Both regions receive extensive innervation directly from serotonergic synapses originating in the dorsal raphe nucleus, which have been shown to play an important role in emotional learning (Lesch & Waider, 2012); because classical psychedelics may play a more direct role in modulating this serotonergic innervation, it is possible that fear conditioning results (in addition to the anxiolytic effects of psychedelics) cannot be attributed to a shift in balance between apical and basal synapses induced by psychedelic administration. We will provide a more detailed review of these results in the text, as well as more clarity regarding their relation to our model.

    Reviewer 1 Concerns:

    Is it reasonable to assign a scalar parameter ‘alpha’ to the effects of classical psychedelics? And is your proposed mechanism of action unique to classical psychedelics? E.g. Could this idea also apply to kappa opioid agonists, ketamine, or the neural mechanisms of hallucination disorders?

    We will clarify that within our model ‘alpha’ is a parameter that reflects the balance between apical and basal synapses in determining the activity of neurons in the network. For the sake of simplicity we used a single ‘alpha’ parameter, but realistically, each neuron would have its own ‘alpha’ parameter, and different layers or individual neurons could be affected differentially by the administration of any particular drug; therefore, our scalar ‘alpha’ value can be thought of as a mean parameter for all neurons, disregarding heterogeneity across individual neurons.

    There are many different mechanisms that could theoretically affect this ‘alpha’ parameter, including: 5-HT2a receptor agonism, kappa opioid receptor binding, ketamine administration, or possibly the effects of genetic mutations underlying the pathophysiology of complex developmental hallucination disorders. We focused exclusively on 5-HT2a receptor agonism for this study because the mechanism is comparatively simple and extensively characterized, but similar mechanisms may well be responsible for the hallucinatory symptoms of a variety of drugs and disorders.

    Can you clarify the role of 5-HT2a receptor expression on interneurons within your model?

    While we mostly focused on the effects of 5-HT2a receptors on the apical dendrites of pyramidal neurons, these receptors are also expressed on soma-targeting parvalbumin (PV) interneurons. This expression on PV interneurons is consistent with our proposed psychedelic mechanism of action, because it could lead to a coordinated decrease in the influence of somatic and proximal dendritic inputs while increasing the influence of apical dendritic inputs. We will elaborate on this point, and will move the discussion earlier in the text.

    Discussions of indigenous use of psychedelics over millenia may amount to over-romanticization.

    We will take great care to conduct a more thorough literature review to reevaluate our statement regarding indigenous psychedelic use (including the citations you suggested), and will either provide a more careful statement or remove this discussion from our introduction entirely, as it has little bearing on the rest of the text. The Ethics Statement will also be modified accordingly.

    You isolate the 5-HT2a agonism as the mechanism of action underlying ‘alpha’ in your model, but there exist 5-HT2a agonists that do not have hallucinatory effects (e.g. lisuride). How do you explain this?

    Lisuride has much-reduced hallucinatory effects compared to other psychedelic drugs at clinical doses (though it does indeed induce hallucinations at high doses; Marona-Lewicka et al., 2002), and we should note that serotonin (5-HT) itself is pervasive in the cortex without inducing hallucinatory effects during natural function. Similarly, MDMA is a partial agonist for 5-HT2a receptors, but it has much-reduced perceptual hallucination effects relative to classical psychedelics (Green et al., 2003) in addition to many other effects not induced by classical psychedelics.

    Therefore, while we argue that 5-HT2a agonism induces an increase in influence of apical dendritic compartments and a decrease in influence of basal/somatic compartments, and that this change induces hallucinations, we also note that there are many other factors that control whether or not hallucinations are ultimately produced, so that not all 5-HT2a agonists are hallucinogenic. We will discuss two such factors in our revision: 5-HT receptor binding affinity and cellular membrane permeability.

    Importantly, many 5-HT2a receptor agonists are also 5-HT1a receptor agonists (e.g. serotonin itself and lisuride), while MDMA has also been shown to increase serotonin, norepinephrine, and dopamine release (Green et al., 2003). While 5-HT2a receptor agonism has been shown to reduce sensory stimulus responses (Michaiel et al., 2019), 5-HT1a receptor agonism inhibits spontaneous cortical activity (Azimi et al., 2020); thus one might expect the net effect of administering serotonin or a nonselective 5-HT receptor agonist to be widespread inhibition of a circuit, as has been observed in visual cortex (Azimi et al., 2020). Therefore, selective 5-HT2a agonism is critical for the induction of hallucinations according to our model, though any intervention that jointly excites pyramidal neurons’ apical dendrites and inhibits their basal/somatic compartments across a broad enough area of cortex would be predicted to have a similar effect. Lisuride has a much higher binding affinity for 5-HT1a receptors than, for instance, LSD (Marona-Lewicka et al., 2002).

    Secondly, it has recently been shown that both the head-twitch effect (a coarse behavioral readout of hallucinations in animals) and the plasticity effects of psychedelics are abolished when administering 5-HT2a agonists that are impermeable to the cellular membrane because of high polarity, and that these effects can be rescued by temporarily rendering the cellular membrane permeable (Vargas et al., 2023). This suggests that the critical hallucinatory effects of psychedelics (apical excitation according to our model) may be mediated by intracellular 5-HT2a receptors. Notably, serotonin itself is not membrane permeable in the cortex.

    Therefore, either of these two properties could play a role in whether a given 5-HT2a agonist induces hallucinatory effects. We will provide a considerably extended discussion of these nuances in our revision.

    Your model proposes that an increase in top-down influence on neural activity underlies the hallucinatory effects of psychedelics. How do you explain experimental results that show increases in bottom-up functional connectivity (either from early sensory areas or the thalamus)?

    Firstly, we should note that our proposed increase in top-down influence is a causal, biophysical property, not necessarily a statistical/correlative one. As such, we will stress that the best way to test our model is via direct intervention in cortical microcircuitry, as opposed to correlative approaches taken by most fMRI studies, which have shown mixed results with regard to this particular question. Correlative approaches can be misleading due to dense recurrent coupling in the system, and due to the coarse temporal and spatial resolution provided by noninvasive recording technologies (changes in statistical/functional connectivity do not necessarily correspond to changes in causal/mechanistic connectivity, i.e. correlation does not imply causation).

    There are two experimental results that appear to contradict our hypothesis that deserve special consideration in our revision. The first shows an increase in directional thalamic influence on the distributed cortical networks after psychedelic administration (Preller et al., 2018). To explain this, we note that this study does not distinguish between lower-order sensory thalamic nuclei (e.g. the lateral and medial geniculate nuclei receiving visual and auditory stimuli respectively) and the higher-order thalamic nuclei that participate in thalamocortical connectivity loops (Whyte et al., 2024). Subsequent more fine-grained studies have noted an increase in influence of higher order thalamic nuclei on the cortex (Pizzi et al., 2023; Gaddis et al., 2022), and in fact extensive causal intervention research has shown that classical psychedelics (and 5-HT2a agonism) decrease the influence of incoming sensory stimuli on the activity of early sensory cortical areas, indicating decoupling from the sensory thalamus (Evarts et al., 1955; Azimi et al., 2020; Michaiel et al. 2019). The increased influence of higher-order thalamic nuclei is consistent with both the cortico-striatal-thalamo-cortical (CTSC) model of psychedelic action as well as the oneirogen hypothesis, since higher-order thalamic inputs modulate the apical dendrites of pyramidal neurons in cortex (Whyte et al., 2024).

    The second experimental result notes that DMT induces traveling waves during resting state activity that propagate from early visual cortex to deeper cortical layers (Alamia et al., 2020). There are several possibilities that could explain this phenomenon: 1) it could be due to the aforementioned difficulties associated with directed functional connectivity analyses, 2) it could be due to a possible high binding affinity for DMT in the visual cortex relative to other brain areas, or 3) it could be due to increases in apical influence on activity caused by local recurrent connectivity within the visual cortex which, in the absence of sensory input, could lead to propagation of neural activity from the visual cortex to the rest of the brain. This last possibility is closest to the model proposed by (Ermentrout & Cowan, 1979), and which we believe would be best explained within our framework by a topographically connected recurrent network architecture trained on video data; a potentially fruitful direction for future research.

    Shouldn’t the hallucinations generated by your model look more ‘psychedelic,’ like those produced by the DeepDream algorithm?

    We believe that the differences in hallucination visualization quality between our algorithm and DeepDream are mostly due to differences in the scale and power of the models used across these two studies. We are confident that with more resources (and potentially theoretical innovations to improve the Wake-Sleep algorithm’s performance) the produced hallucination visualizations could become more realistic, but we believe this falls outside the scope of the present study.

    We note that more powerful generative models trained with backpropagation are able to produce surreal images of comparable quality (Rezende et al., 2014; Goodfellow et al., 2020; Vahdat & Kautz, 2020), though these have not yet been used as a model of psychedelic hallucinations. However, the DeepDream model operates on top of large pretrained image processing models, and does not provide a biologically mechanistic/testable interpretation of its hallucination effects. When training smaller models with a local synaptic plasticity rule (as opposed to backpropagation), the hallucination effects are less visually striking due to the reduced quality of our trained generative model, though they are still strongly tied to the statistics of sensory inputs, as quantified by our correlation similarity metric (Fig. 5b). We will provide a more detailed explanation of this phenomenon when we discuss our model limitations in our revised manuscript.

    Your model assumes domination by entirely bottom-up activity during the ‘wake’ phase, and domination entirely by top-down activity during ‘sleep,’ despite experimental evidence indicating that a mixture of top-down and bottom-up inputs influence neural activity during both stages in the brain. How do you explain this?

    Our use of the Wake-Sleep algorithm, in which top-down inputs (Sleep) or bottom-up inputs (Wake) dominate network activity is an over-simplification made within our model for computational and theoretical reasons. Models that receive a mixture of top-down and bottom-up inputs during ‘Wake’ activity do exist (in particular the closely related Boltzmann machine (Ackley et al., 1985)), but these models are considerably more computationally costly to train due to a need to run extensive recurrent network relaxation dynamics for each input stimulus. Further, these models do not generalize as cleanly to processing temporal inputs. For this reason, we focused on the Wake-Sleep algorithm, at the cost of some biological realism, though we note that our model should certainly be extended to support mixed apical-basal waking regimes. We will make sure to discuss this in our ‘Model Limitations’ section.

    Your model proposes that 5-HT2a agonism enhances glutamatergic transmission, but this is not true in the hippocampus, which shows decreases in glutamate after psychedelic administration.

    We should note that our model suggests only compartment specific increases in glutamatergic transmission; as such, our model does not predict any particular directionality for measures of glutamatergic transmission that includes signaling at both apical and basal compartments in aggregate, as was measured in the provided study (Mason et al., 2020).

    You claim that your model is consistent with the Entropic Brain theory, but you report increases in variance, not entropy. In fact, it has been shown that variance decreases while entropy increases under psychedelic administration. How do you explain this discrepancy?

    Unfortunately, ‘entropy’ and ‘variance’ are heavily overloaded terms in the noninvasive imaging literature, and the particularities of the method employed can exert a strong influence on the reported effects. The reduction in variance reported by (Carhart-Harris et al., 2016) is a very particular measure: they are reporting the variance of resting state synchronous activity, averaged across a functional subnetwork that spans many voxels; as such, the reduction in variance in this case is a reduction in broad, synchronous activity. We do not have any resting state synchronous activity in our network due to the simplified nature of our model (particularly an absence of recurrent temporal dynamics), so we see no reduction in variance in our model due to these effects.

    Other studies estimate ‘entropy’ or network state disorder via three different methods that we have been able to identify. 1) (Carhart-Harris et al., 2014) uses a different measure of variance: in this case, they subtract out synchronous activity within functional subnetworks, and calculate variability across units in the network. This measure reports increases in variance (Fig. 6), and is the closest measure to the one we employ in this study. 2) (Lebedev et al., 2016) uses sample entropy, which is a measure of temporal sequence predictability. It is specifically designed to disregard highly predictable signals, and so one might imagine that it is a measure that is robust to shared synchronous activity (e.g. resting state oscillations). 3) (Mediano et al., 2024) uses Lempel-Ziv complexity, which is, similar to sample entropy, a measure of sequence diversity; in this case the signal is binarized before calculation, which makes this method considerably different from ours. All three of the preceding methods report increases in sequence diversity, in agreement with our quantification method. Our strongest explanation for why the variance calculation in (Carhart-Harris et al., 2016) produces a variance reduction is therefore due to a reduction in low-rank synchronous activity in subnetworks during resting state.

    As for whether the entropy increase is meaningful: we share Reviewer 1’s concern that increases in entropy could simply be due to a higher degree of cognitive engagement during resting state recordings, due to the presence of sensory hallucinations or due to an inability to fall asleep. This could explain why entropy increases are much more minimal relative to non-hallucinating conditions during audiovisual task performance (Siegel et al., 2024; Mediano et al., 2024). However, we can say that our model is consistent with the Entropic Brain Theory without including any form of ‘cognitive processing’: we observe increases in variability during resting state in our model, but we observe highly similar distributions of activity when averaging over a wide variety of sensory stimulus presentations (Fig. 5b-c). This is because variability in our model is not due to unstructured noise: it corresponds to an exploration of network states that would ordinarily be visited by some stimulus. Therefore, when averaging across a wide variety of stimuli, the distribution of network states under hallucinating or non-hallucinating conditions should be highly similar.

    One final point of clarification: here we are distinguishing Entropic Brain Theory from the REBUS model–the oneirogen hypothesis is consistent with the increase in entropy observed experimentally, but in our model this entropy increase is not due to increased influence of bottom-up inputs (it is due instead to an increase in top-down influence). Therefore, one could view the oneirogen hypothesis as consistent with EBT, but inconsistent with REBUS.

    You relate your plasticity rule to behavioral-timescale plasticity (BTSP) in the hippocampus, but plasticity has been shown to be reduced in the hippocampus after psychedelic administration. Could you elaborate on this connection?

    When we were establishing a connection between our ‘Wake-Sleep’ plasticity rule and BTSP learning, the intended connection was exclusively to the mathematical form of the plasticity rule, in which activity in the apical dendrites of pyramidal neurons functions as an instructive signal for plasticity in basal synapses (and vice versa): we will clarify this in the text. Similarly, we point out that such a plasticity rule tends to result in correlated tuning between apical and basal dendritic compartments, which has been observed in hippocampus and cortex: this is intended as a sanity check of our mapping of the Wake-Sleep algorithm to cortical microcircuitry, and has limited further bearing on the effects of psychedelics specifically.

    Reduction in plasticity in the hippocampus after psychedelic administration could be due to a complementary learning systems-type model, in which the hippocampus becomes partly decoupled from the cortex during REM sleep (Singh et al., 2022); were this to be the case, it would not be incompatible with our model, which is mostly focused on the cortex. Notably, potentiating 5HT-2a receptors in the ventral hippocampus does not induce the head-twitch response, though it does produce anxiolytic effects (Tiwari et al., 2024), indicating that the hallucinatory and anxiolytic effects of classical psychedelics may be partly decoupled.

    Reviewer 2 Concerns:

    Could you provide visualizations of the ‘ripple’ phenomenon that you’re referring to?

    We will do this! For now, you can get a decent understanding of what the ‘ripple effect’ looks like from the ‘eyes closed’ hallucination condition for networks trained on CIFAR10 (Fig. 2d). The ripple effect that we are referring to is very similar, except it is superimposed on a naturalistic image under ordinary viewing conditions; to give a higher quality visualization of the ripple phenomenon itself, we will subtract out the static contribution of the image itself, leaving only the ripple phenomenon.

    Could you provide a more nuanced description of alternative roles for top-down feedback, beyond being used exclusively for learning as depicted in your model?

    For the sake of simplicity, we only treat top-down inputs in our model as a source of an instructive teaching signal, the originator of generative replay events during the Sleep phase, and as the mechanism of hallucination generation. However, as discussed in a response to a previous question, in the cortex pyramidal neurons receive and respond to a mixture of top-down and bottom-up processing.

    There are a variety of theories for what role top-down inputs could play in determining network activity. To name several, top-down input could function as: 1) a denoising/pattern completion signal (Kadkhodaie & Simoncelli, 2021), 2) a feedback control signal (Podlaski & Machens, 2020), 3) an attention signal (Lindsay, 2020), 4) ordinary inputs for dynamic recurrent processing that play no specialized role distinct from bottom-up or lateral inputs except to provide inputs from higher-order association areas or other sensory modalities (Kar et al., 2019; Tugsbayar et al., 2025). Though our model does not include these features, they are perfectly consistent with our approach.

    In particular, denoising/pattern completion signals in the predictive coding framework (closely related to the Wake-Sleep algorithm) also play a role as an instructive learning signal (Salvatori et al., 2021); and top-down control signals can play a similar role in some models (Gilra & Gerstner, 2017; Meulemans et al., 2021). Thus, options 1 and 2 are heavily overlapping with our approach, and are a natural consequence of many biologically plausible learning algorithms that minimize a variational free energy loss (Rao & Ballard, 1997; Ackley et al., 1985). Similarly, top-down attentional signals can exist alongside top-down learning signals, and some models have argued that such signals can be heavily overlapping or mutually interchangeable (Roelfsema & van Ooyen, 2005). Lastly, generic recurrent connectivity (from any source) can be incorporated into the Wake-Sleep algorithm (Dayan & Hinton, 1996), though we avoided doing this in the present study due to an absence of empirical architecture exploration in the literature and the computational complexity associated with training on time series data.

    To conclude, there are a variety of alternative functions proposed for top-down inputs onto pyramidal neurons in the cortex, and we view these additional features as mutually compatible with our approach; for simplicity we did not include them in our model, but we believe that these features are unlikely to interfere with our testable predictions or empirical results.

  5. Version published to 10.7554/elife.105968.1 on eLife
  6. Version published to 10.7554/elife.105968 on eLife
  7. Version published to 10.1101/2024.09.27.615483v2 on bioRxiv
  8. Version published to 10.1101/2024.09.27.615483v1 on bioRxiv