Search

Language is not only used to transmit neutral information; we often seek to persuade by arguing in favor of a particular view. Persuasion raises a number of challenges for classical accounts of belief updating, as information cannot be taken at face value. How should listeners account for a speaker's "hidden agenda" when incorporating new information? Here, we extend recent probabilistic models of recursive social reasoning to allow for persuasive goals and show that our model provides a pragmatic account for why weakly favorable arguments may backfire, a phenomenon known as the weak evidence effect. Critically, this model predicts a systematic relationship between belief updates and expectations about the information source: weak evidence should only backfire when speakers are expected to act under persuasive goals and prefer the strongest evidence. We introduce a simple experimental paradigm called the Stick Contest to measure the extent to which the weak evidence effect depends on speaker expectations, and show that a pragmatic listener model accounts for the empirical data better than alternative models. Our findings suggest further avenues for rational models of social reasoning to illuminate classical decision-making phenomena., Comment: in press at Open Mind

Piecing together the meaning of a narrative requires understanding both individual words and the intricate relationships between them. How does the brain construct this kind of rich, contextual meaning? Recently, a new class of artificial neural networks—based on the Transformer architecture—has revolutionized the field of language modeling. Transformers integrate information across words via multiple layers of structured circuit computations, forming increasingly contextualized representations of linguistic content. In this paper, we deconstruct these circuit computations and analyze the associated “transformations” (alongside the more commonly studied “embeddings”) to provide a fine-grained window onto linguistic computations in the human brain. Using functional MRI data acquired while participants listened to naturalistic spoken stories, we find that these transformations capture a hierarchy of linguistic computations across cortex, with transformations at later layers in the model mapping onto higher-level language areas in the brain. We then decompose these transformations into individual, functionally-specialized “attention heads” and demonstrate that the emergent syntactic computations performed by individual heads correlate with predictions of brain activity in specific cortical regions. These heads fall along gradients corresponding to different layers, contextual distances, and syntactic dependencies in a low-dimensional cortical space. Our findings indicate that large language models and the cortical language network converge on similar trends of computational specialization for processing natural language.

Languages are powerful solutions to coordination problems: they provide stable, shared expectations about how the words we say correspond to the beliefs and intentions in our heads. Yet language use in a variable and non-stationary social environment requires linguistic representations to be flexible: old words acquire new ad hoc or partner-specific meanings on the fly. In this paper, we introduce CHAI (Continual Hierarchical Adaptation through Inference), a hierarchical Bayesian theory of coordination and convention formation that aims to reconcile the long-standing tension between these two basic observations. We argue that the central computational problem of communication is not simply transmission, as in classical formulations, but continual learning and adaptation over multiple timescales. Partner-specific common ground quickly emerges from social inferences within dyadic interactions, while community-wide social conventions are stable priors that have been abstracted away from interactions with multiple partners. We present new empirical data alongside simulations showing how our model provides a computational foundation for several phenomena that have posed a challenge for previous accounts: (1) the convergence to more efficient referring expressions across repeated interaction with the same partner, (2) the gradual transfer of partner-specific common ground to strangers, and (3) the influence of communicative context on which conventions eventually form., Comment: In press at Psychological Review

People use a wide range of communicative acts across different modalities, from concrete demonstrations to abstract language. While these modalities are typically studied independently, we take a comparative approach and ask when and why one modality might outperform another. We present a series of real-time, multi-player experiments asking participants to teach concepts using either demonstrations or language. Our first experiment (N=416) asks when language might outperform demonstration. We manipulate the complexity of the concept being taught and find that language communicates complex concepts more effectively than demonstration. We then ask why language succeeds in this setting. We hypothesized that language allowed teachers to reference abstract object features (e.g., shapes and colors), while demonstration teachers could only provide concrete examples (specific positive or negative objects). To test this hypothesis, our second experiment (N=568) ablated object features from the teacher's interface. This manipulation severely impaired linguistic (but not demonstrative) teaching. Our findings suggest that language communicates complex concepts by directly transmitting abstract rules. In contrast, demonstrations transmit examples, requiring the learner to infer the rules.

Two fundamental issues in memory research concern when later experiences strengthen or weaken initial memories and when the two memories become linked or remain independent. A promising candidate for explaining these issues is semantic relatedness. Here, across five paired-associate learning experiments (N=1000), we systematically varied the semantic relatedness between initial and later cues, initial and later targets, or both. We found that learning retroactively benefited long-term memory performance for semantically related words (vs. unshown control words), and these benefits increased as a function of relatedness. Critically, memory dependence between initial and later pairs also increased with relatedness, suggesting that pre-existing semantic relationships promote interdependence for memories formed across episodes. We also found that modest retroactive benefits, but not interdependencies, emerged when subjects learned via studying rather than practice testing. These findings demonstrate that semantic relatedness during new learning retroactively strengthens old associations while scaffolding new ones into well-fortified memory traces.

There is substantial variability in the expectations that communication partners bring into interactions, creating the potential for misunderstandings. To directly probe these gaps and our ability to overcome them, we propose a communication task based on color-concept associations. In Experiment 1, we establish several key properties of the mental representations of these expectations, or lexical priors, based on recent probabilistic theories. Associations are more variable for abstract concepts, variability is represented as uncertainty within each individual, and uncertainty enables accurate predictions about whether others are likely to share the same association. In Experiment 2, we then examine the downstream consequences of these representations for communication. Accuracy is initially low when communicating about concepts with more variable associations, but rapidly increases as participants form ad hoc conventions. Together, our findings suggest that people cope with variability by maintaining well-calibrated uncertainty about their partner and appropriately adaptable representations of their own., in press at Cognition

We explore unconstrained natural language feedback as a learning signal for artificial agents. Humans use rich and varied language to teach, yet most prior work on interactive learning from language assumes a particular form of input (e.g., commands). We propose a general framework which does not make this assumption, using aspect-based sentiment analysis to decompose feedback into sentiment about the features of a Markov decision process. We then perform an analogue of inverse reinforcement learning, regressing the sentiment on the features to infer the teacher's latent reward function. To evaluate our approach, we first collect a corpus of teaching behavior in a cooperative task where both teacher and learner are human. We implement three artificial learners: sentiment-based "literal" and "pragmatic" models, and an inference network trained end-to-end to predict latent rewards. We then repeat our initial experiment and pair them with human teachers. All three successfully learn from interactive human feedback. The sentiment models outperform the inference network, with the "pragmatic" model approaching human performance. Our work thus provides insight into the information structure of naturalistic linguistic feedback as well as methods to leverage it for reinforcement learning., 9 pages, 4 figures. AAAI '21

Long-term but not short-term memory and synaptic plasticity in many brain areas require neurotrophin signaling, transcription, and epigenetic mechanisms including DNA methylation. However, it has been difficult to relate these cellular mechanisms directly to behavior because of the immense complexity of the mammalian brain. To address that problem, we and others have examined numerically simpler systems such as the hermaphroditic marine mollusk Aplysia californica. As a further simplification, we have used a semi-intact preparation of the Aplysia siphon withdrawal reflex in which it is possible to relate cellular plasticity directly to behavioral learning. We find that inhibitors of neurotrophin signaling, transcription, and DNA methylation block sensitization and classical conditioning beginning ∼1 h after the start of training, which is in the time range of an intermediate-term stage of plasticity that combines elements of short- and long-term plasticity and may form a bridge between them. Injection of decitabine (an inhibitor of DNA methylation that may have other actions in these experiments) into an LE sensory neuron blocks the neural correlates of conditioning in the same time range. In addition, we found that both DNA and RNA methylation in the abdominal ganglion are correlated with learning in the same preparations. These results begin to suggest the functions and integration of these different molecular mechanisms during behavioral learning.

Learning and memory have long been thought to involve changes in synaptic connections between neurons. However, in many cases learning-related plasticity also involves changes in the excitability of neurons. These findings have raised questions about the relative importance of these two types of mechanisms to behavioral learning, and also about the extent to which they involve shared or unique molecular mechanisms. We have taken a reductionist approach to these questions by addressing them in a simple model organism, Aplysia californica. Studies of a semi-intact Aplysia siphon withdrawal preparation suggest that classical conditioning involves an increase in the evoked firing of sensory neurons (SNs) as well as facilitation of the monosynaptic PSP to motor neurons (MNs). Furthermore, these two mechanisms may act cooperatively at the cellular level: increased SN firing produces more PSPs, each of which is facilitated, leading to a multiplicative increase in depolarization of the MN and siphon withdrawal. The changes in SN firing and the monosynaptic PSP also share several mechanisms at the molecular level, suggesting that they may both be due in part to a decrease in K(+) current that causes an increase in SN excitability as well as an increase in SN spike width and thus increased transmitter release. However, changes in the monosynaptic PSP also involve additional mechanisms that are not shared and may affect different aspects of synaptic transmission as well. Studies of operant conditioning of feeding suggest that it involves similar mechanisms as classical conditioning of siphon withdrawal. In particular, for both types of associative learning adenylyl cyclase appears to serve as a molecular coincidence detector that leads to increased activation of PKA and changes in excitability of key neurons in the neural circuit. Furthermore, in both cases those changes in excitability make an important contribution to the behavioral learning.

One of the major questions in psychology is whether associative and nonassociative learning are fundamentally different or whether they involve similar processes and mechanisms. We have addressed this question by comparing mechanisms of a nonassociative form of learning, sensitization, and an associative form of learning, classical conditioning of the siphon-withdrawal reflex of hermaphroditic Aplysia. In an analog of differential conditioning, action potentials in one siphon sensory neuron (SN) were paired with shock to the pedal nerves, producing activity-dependent presynaptic facilitation, and action potentials in another SN were unpaired with the shock as a control. The difference between paired and unpaired training is a measure of associative plasticity. Before and after this training, we voltage clamped each SN and measured the outward current during depolarizing pulses. There was a significantly greater decrease in the net outward current in the paired SN than in the unpaired SN. We obtained similar results when we substituted the depolarizing voltage clamp pulse for action potentials during training. We then bathed the ganglion in serotonin as a measure of nonassociative plasticity. The current that was modulated differentially (paired−unpaired) had time and voltage dependencies similar to the current that was modulated by serotonin (Is). These results suggest that an associative form of plasticity, activity-dependent presynaptic facilitation underlying conditioning, involves enhanced modulation of the same ionic current as a nonassociative form, normal presynaptic facilitation underlying sensitization.

Hyperpolarization-activated, cyclic nucleotide-gated cation (HCN) channels are critical regulators of neuronal excitability, but less is known about their possible roles in synaptic plasticity and memory circuits. Here, we characterized the HCN gene organization, channel properties, distribution, and involvement in associative and nonassociative forms of learning in Aplysia californica. Aplysia has only one HCN gene, which codes for a channel that has many similarities to the mammalian HCN channel. The cloned acHCN gene was expressed in Xenopus oocytes, which displayed a hyperpolarization-induced inward current that was enhanced by cGMP as well as cAMP. Similarly to its homologs in other animals, acHCN is permeable to K(+) and Na(+) ions, and is selectively blocked by Cs(+) and ZD7288. We found that acHCN is predominantly expressed in inter- and motor neurons, including LFS siphon motor neurons, and therefore tested whether HCN channels are involved in simple forms of learning of the siphon-withdrawal reflex in a semiintact preparation. ZD7288 (100 μM) significantly reduced an associative form of learning (classical conditioning) but had no effect on two nonassociative forms of learning (intermediate-term sensitization and unpaired training) or baseline responses. The HCN current is enhanced by nitric oxide (NO), which may explain the postsynaptic role of NO during conditioning. HCN current in turn enhances the NMDA-like current in the motor neurons, suggesting that HCN channels contribute to conditioning through this pathway.

Whereas short-term plasticity is often initiated on one side of the synapse, long-term plasticity involves coordinated changes on both sides, implying extracellular signaling. We have investigated the possible signaling role of an Aplysia neurotrophin (ApNT) in facilitation induced by serotonin (5HT) at sensory-to-motor neuron synapses in culture. ApNT is an ortholog of mammalian BDNF, which has been reported to act as either an anterograde, retrograde, or autocrine signal, so that its pre- and postsynaptic sources and targets remain unclear. We now report that ApNT acts as a presynaptic autocrine signal that forms part of a positive feedback loop with ApTrk and PKA. That loop stimulates spontaneous transmitter release, which recruits postsynaptic mechanisms, and presynaptic protein synthesis during the transition from short- to intermediate-term facilitation and may also initiate gene regulation to trigger the transition to long-term facilitation. These results suggest that a presynaptic ApNT feedback loop plays several key roles during consolidation of learning-related synaptic plasticity.

Whereas short-term synaptic plasticity is often either pre- or postsynaptic, intermediate- and long-term plasticity generally require coordinated pre- and postsynaptic mechanisms. Thus, the transition from presynaptic short-term facilitation (STF) to intermediate-term facilitation (ITF) induced by 5HT at Aplysia sensory-to-motor neuron synapses requires the recruitment of postsynaptic mechanisms and activation of protein synthesis in both neurons. In the companion paper to this report, we found that presynaptic autocrine signaling by an Aplysia neurotrophin (ApNT) forms a positive feedback loop that drives the synapses from STF to ITF. Here we report that ApNT also acts through both anterograde and retrograde signaling to form a transsynaptic positive feedback loop that orchestrates cellular functions in both the presynaptic and postsynaptic neurons during the induction of ITF. These two feedback loops activate protein synthesis in each synaptic compartment, which in both cases depends on signaling from the other synaptic compartment. These results suggest that the pre- and postsynaptic compartments act as one functional unit during the consolidation of learning-related facilitation induced by 5HT.

Whereas short-term synaptic plasticity is often either pre- or postsynaptic, intermediate- and long-term plasticity generally require coordinated pre- and postsynaptic mechanisms. Thus, the transition from presynaptic short-term facilitation (STF) to intermediate-term facilitation (ITF) induced by 5HT at Aplysia sensory-to-motor neuron synapses requires the recruitment of postsynaptic mechanisms and activation of protein synthesis in both neurons. In the companion paper to this report, we found that presynaptic autocrine signaling by an Aplysia neurotrophin (ApNT) forms a positive feedback loop that drives the synapses from STF to ITF. Here we report that ApNT also acts through both anterograde and retrograde signaling to form a transsynaptic positive feedback loop that orchestrates cellular functions in both the presynaptic and postsynaptic neurons during the induction of ITF. These two feedback loops activate protein synthesis in each synaptic compartment, which in both cases depends on signaling from the other synaptic compartment. These results suggest that the pre- and postsynaptic compartments act as one functional unit during the consolidation of learning-related facilitation induced by 5HT.

The gill- and siphon-withdrawal reflex of Aplysia is a simple defensive behavior that is mediated in part by a monosynaptic pathway. Despite this remarkable simplicity, the reflex can be modified by several forms of learning, including habituation, dishabituation, sensitization, and classical and operant conditioning. The forms of learning that have been explored in research studies exhibit many of the behavioral properties of learning in mammals, suggesting that they may involve similar neuronal mechanisms. In 1984, we proposed cellular mechanisms for several higher-order features of conditioning and incorporated those ideas in a quantitative model that simulated a broad range of behavioral properties. In this chapter, we summarize the current body of knowledge about the behavior, circuitry, and cellular and molecular mechanisms of learning and memory of the reflex. We then review our original model and suggest how recent advances may explain some of the behavioral properties that this model could not.

Cellular studies of memory suggest that experience-dependent synaptic plasticity is a fundamental mechanism by which memories are encoded, processed, and stored within the brain. In this chapter, we focus on recent advances in our understanding of the molecular mechanisms of a simple form of learning-related plasticity, presynaptic facilitation, that contributes to short-term, intermediate-term, and long-term forms of memory in the marine invertebrate Aplysia californica . We also briefly review mechanisms of presynaptic plasticity in other experimental systems, which share many similarities with those in Aplysia and may play important roles in memory as well.

Studies of implicit and explicit memory suggest that modulation of synaptic strength and structure is a fundamental mechanism by which memories are encoded, processed, and stored within the brain. Two model systems have been extensively studied as examples of these two forms of memory: sensitization in the marine snail Aplysia californica as an example of implicit memory and spatial memory formation in rodents as an example of explicit memory. In this chapter, we discuss and compare critical synaptic sites and cellular and molecular mechanisms of implicit and explicit memory storage, and describe new approaches to relating these mechanisms to behavior.

Recent studies in Aplysia have identified a new variation of synaptic plasticity in which modulatory transmitters enhance spontaneous release of glutamate, which then acts on postsynaptic receptors to recruit mechanisms of intermediate- and long-term plasticity. In this review I suggest the hypothesis that similar plasticity occurs in mammals, where it may contribute to reward, memory, and their dysfunctions in several psychiatric disorders. In Aplysia, spontaneous release is enhanced by activation of presynaptic serotonin receptors, but presynaptic D1 dopamine receptors or nicotinic acetylcholine receptors could play a similar role in mammals. Those receptors enhance spontaneous release of glutamate in hippocampus, entorhinal cortex, prefrontal cortex, ventral tegmental area, and nucleus accumbens. In all of those brain areas, glutamate can activate postsynaptic receptors to elevate Ca2+ and engage mechanisms of early-phase long-term potentiation (LTP), including AMPA receptor insertion, and of late-phase LTP, including protein synthesis and growth. Thus, presynaptic receptors and spontaneous release may contribute to postsynaptic mechanisms of plasticity in brain regions involved in reward and memory, and could play roles in disorders that affect plasticity in those regions, including addiction, Alzheimer’s disease, schizophrenia, and attention deficit hyperactivity disorder (ADHD).

Imaging studies have shown that even the earliest phases of long-term plasticity are accompanied by the rapid recruitment of synaptic components, which generally requires actin polymerization and may be one of the first steps in a program that can lead to the formation of new stable synapses during late-phase plasticity. However, most of those results come from studies of long-term potentiation in rodent hippocampus and might not generalize to other forms of synaptic plasticity or plasticity in other brain areas and species. For example, recruitment of presynaptic proteins during long-term facilitation by 5HT in Aplysia is delayed for several hours, suggesting that whereas activity-dependent forms of plasticity, such as long-term potentiation, involve rapid recruitment of presynaptic proteins, neuromodulatory forms of plasticity, such as facilitation by 5HT, involve more delayed recruitment. To begin to explore this hypothesis, we examined an activity-dependent form of plasticity, homosynaptic potentiation produced by tetanic stimulation of the presynaptic neuron in Aplysia . We found that homosynaptic potentiation involves presynaptic but not postsynaptic actin and a rapid (under 10 min) increase in the number of clusters of the presynaptic vesicle-associated protein synaptophysin. These results indicate that rapid recruitment of synaptic components is not limited to hippocampal potentiation and support the hypothesis that activity-dependent types of plasticity involve rapid recruitment of presynaptic proteins, whereas neuromodulatory types of plasticity involve more delayed recruitment.

A good deal is known about psychological factors that contribute to reward value, but very little is known about how reward is computed in the brain. Integration in the circuit for brain stimulation reward may provide a simple model system. Parametric studies have favored the idea that the integration is linear, although a few have suggested that it involves adaptation. However, these studies have generally not focused on relevance to psychological factors such as temporal discounting of reward. To begin to address these issues, I gave rats a choice between a simple train of 2 seconds of stimulation at a uniform frequency and a compound train of 1 second of stimulation at either 50, 150, or 250 Hz immediately followed by 1 second of stimulation at 50, 150, or 250 Hz, for a total of nine compound trains. For each compound train, the frequency of the simple train was varied until a level was found that produced indifference. The order of the frequencies was important, with Hi-Lo compound stimulation having a higher matching frequency than the symmetrical Lo-Hi. Furthermore, there was an interaction between the first and second frequencies, with the second frequency counting less when the first frequency was high. The results could be fit by computational models incorporating the ideas of either (a) adaptation or (b) temporal discounting and a power function psychological scale. These results suggest that integration in the brain stimulation reward system is nonlinear, and involves either adaptation or temporal discounting of reward value.

Synaptic plasticity and learning involve different mechanisms depending on the following: (1) the stage of plasticity and (2) the history of plasticity, or metaplasticity. However, little is known about how these two factors are related. We have addressed that question by examining mechanisms of synaptic plasticity during short-term and intermediate-term behavioral sensitization and dishabituation in a semi-intact preparation of theAplysiasiphon-withdrawal reflex. Dishabituation differs from sensitization in that it is preceded by habituation, and is thus a paradigm for metaplasticity. We find that whereas facilitation during short-term sensitization by one tail shock involves presynaptic covalent modifications by protein kinase A (PKA) and CamKII, facilitation during intermediate-term sensitization by four shocks involves both presynaptic (PKA, CaMKII) and postsynaptic (Ca2+, CaMKII) covalent modifications, as well as both presynaptic and postsynaptic protein synthesis. The facilitation also involves presynaptic spike broadening 2.5 min after either one or four shocks, but not at later times. Dishabituation by four shocks differs from sensitization in several ways. First, it does not involve PKA or CaMKII, but rather involves presynaptic PKC. In addition, unlike sensitization with the same shock, dishabituation by four shocks does not involve protein synthesis or presynaptic spike broadening, and it also does not involve postsynaptic Ca2+. These results demonstrate that not only the mechanisms but also the site of plasticity depend on both the stage of plasticity and metaplasticity during memory formation.

SummaryTranssynaptic interactions between neurons are essential during both developmental and learning-related synaptic growth. We have used Aplysia neuronal cultures to examine the contribution of transsynaptic signals in both types of synapse formation. We find that during de novo synaptogenesis, specific presynaptic innervation is required for the clustering of postsynaptic AMPA-like but not NMDA-like receptors. We further find that the cell adhesion molecule Dscam is involved in these transsynaptic interactions. Inhibition of Dscam either pre- or postsynaptically abolishes the emergence of synaptic transmission and the clustering of AMPA-like receptors. Remodeling of both AMPA-like and NMDA-like receptors also occurs during learning-related synapse formation and again requires the reactivation of Dscam-mediated transsynaptic interactions. Taken together, these findings suggest that learning-induced synapse formation recapitulates, at least in part, aspects of the mechanisms that govern de novo synaptogenesis.

This work reviews research on neural mechanisms of two types of associative learning in the marine mollusk Aplysia, classical conditioning of the gill- and siphon-withdrawal reflex and operant conditioning of feeding behavior. Basic classical conditioning is caused in part by activity-dependent facilitation at sensory neuron-motor neuron (SN-MN) synapses and involves a hybrid combination of activity-dependent presynaptic facilitation and Hebbian potentiation, which are coordinated by trans-synaptic signaling. Classical conditioning also shows several higher-order features, which might be explained by the known circuit connections in Aplysia. Operant conditioning is caused in part by a different type of mechanism, an intrinsic increase in excitability of an identified neuron in the central pattern generator (CPG) for feeding. However, for both classical and operant conditioning, adenylyl cyclase is a molecular site of convergence of the two signals that are associated. Learning in other invertebrate preparations also involves many of the same mechanisms, which may contribute to learning in vertebrates as well.

The simplicity and tractability of the neural circuits mediating behaviors in invertebrates have facilitated the cellular/molecular dissection of neural mechanisms underlying learning. The review has a particular focus on the general principles that have emerged from analyses of an example of nonassociative learning, sensitization in the marine mollusk Aplysia. Learning and memory rely on multiple mechanisms of plasticity at multiple sites of the neuronal circuits, with the relative contribution to memory of the different sites varying as a function of the extent of training and time after training. The same intracellular signaling cascades that induce short-term modifications in synaptic transmission can also be used to induce long-term changes. Although short-term memory relies on covalent modifications of preexisting proteins, long-term memory also requires regulated gene transcription and translation. Maintenance of long-term cellular memory involves both intracellular and extracellular feedback loops, which sustain the regulation of gene expression and the modification of targeted molecules.

Cellular studies of implicit and explicit memory suggest that experience-dependent modulation of synaptic strength and structure is a fundamental mechanism by which these memories are encoded, processed, and stored within the brain. In this review, we focus on recent advances in our understanding of the molecular mechanisms that underlie short-term, intermediate-term, and long-term forms of implicit memory in the marine invertebrate Aplysia cali- fornica, and consider how the conservation of common elements in each form may contribute to the different tem- poral phases of memory storage.

Dishabituation has been thought to be due either to reversal of the process of habituation or to a second process equivalent to sensitization superimposed on habituation. One way to address this question is by testing whether dishabituation and sensitization can be dissociated. Previous studies using this approach in Aplysia have come to different conclusions about the nature of dishabituation, perhaps because those studies differed in many respects, including (1) whether they also observed transient behavioral inhibition, and (2) whether they used test stimuli that activated the LE siphon sensory neurons or as yet unidentified sensory neurons. To attempt to resolve the apparent contradictions between the previous studies, we have explored the importance of these two factors by performing a parametric study of dishabituation and sensitization of gill withdrawal in a simplified preparation that does not exhibit transient behavioral inhibition, using two different test stimuli that are known to activate the LE (Touch) or unidentified (Not Touch) sensory neurons. We find that dishabituation and sensitization in this preparation have similar time courses and generally similar functions of shock intensity. However, under one condition, with the Not Touch stimulus 2.5 min after the shock, dishabituation has a reverse effect of shock intensity. Additional analyses suggest that dishabituation with the Not Touch stimulus 2.5 min after the shock is due to reversal of habituation, whereas 12.5 min after the shock, dishabituation is due to superimposed sensitization. These results thus suggest that dishabituation may involve either process in the same preparation, and begin to define the conditions that favor one or the other.

Long-term potentiation (LTP) in the Schaffer collateral pathway from the CA3 to the CA1 region of the hippocampus is thought to involve postsynaptic mechanisms including Ca 2+ - and CamKII-dependent α-amino-3-hydroxy-5-methyl-4-isoxazolepropionic acid (AMPA) receptor insertion. However, very little is known about possible presynaptic mechanisms. It is easier to address that question at synapses between individual neurons in the CA3 region, where both sides of the synapses are accessible to substances injected into the cell bodies. Previous studies using that method showed that CA3–CA3 LTP involves presynaptic protein kinases as well as postsynaptic receptor insertion. We have extended those findings by exploring the pre- and postsynaptic roles of Ca 2+ and CamKII, and we have also compared results with two induction protocols, 1-Hz-paired and θ-burst-paired, which may involve pre- and/or postsynaptic mechanisms in addition to receptor insertion in CA1. Similar to results in CA1, we find that CA3–CA3 LTP completely depends on postsynaptic Ca 2+ with the 1-Hz-paired protocol but depends only partially on postsynaptic Ca 2+ or CamKII with the θ-burst-paired protocol. Potentiation with that protocol also partially depends on presynaptic Ca 2+ or CamKII, suggesting that the additional mechanisms of potentiation, at least in part, are presynaptic. Furthermore, the pre- and postsynaptic mechanisms seem to act in series, suggesting coordinate regulation of the two sides of the synapses. CA3–CA3 LTP with the 1-Hz-paired protocol also partially depends on presynaptic Ca 2+ , suggesting that it may involve presynaptic mechanisms as well.