Recent studies, conducted almost exclusively in primates, have shown that several cortical areas usually associated with modality-specific sensory processing are subject to influences from other senses. Here we demonstrate using single-unit recordings and estimates of mutual information that visual stimuli can influence the activity of units in the auditory cortex of anesthetized ferrets. In many cases, these units were also acoustically responsive and frequently transmitted more information in their spike discharge patterns in response to paired visual–auditory stimulation than when either modality was presented by itself. For each stimulus, this information was conveyed by a combination of spike count and spike timing. Even in primary auditory areas (primary auditory cortex [A1] and anterior auditory field [AAF]), ∼15% of recorded units were found to have nonauditory input. This proportion increased in the higher level fields that lie ventral to A1/AAF and was highest in the anterior ventral field, where nearly 50% of the units were found to be responsive to visual stimuli only and a further quarter to both visual and auditory stimuli. Within each field, the pure-tone response properties of neurons sensitive to visual stimuli did not differ in any systematic way from those of visually unresponsive neurons. Neural tracer injections revealed direct inputs from visual cortex into auditory cortex, indicating a potential source of origin for the visual responses. Primary visual cortex projects sparsely to A1, whereas higher visual areas innervate auditory areas in a field-specific manner. These data indicate that multisensory convergence and integration are features common to all auditory cortical areas but are especially prevalent in higher areas.

We characterized the functional organization of different fields within the auditory cortex of anaesthetized ferrets. As previously reported, the primary auditory cortex, A1, and the anterior auditory field, AAF, are located on the middle ectosylvian gyrus. These areas exhibited a similar tonotopic organization, with high frequencies represented at the dorsal tip of the gyrus and low frequencies more ventrally, but differed in that AAF neurons had shorter response latencies than those in A1. On the basis of differences in frequency selectivity, temporal response properties and thresholds, we identified four more, previously undescribed fields. Two of these are located on the posterior ectosylvian gyrus and were tonotopically organized. Neurons in these areas responded robustly to tones, but had longer latencies, more sustained responses and a higher incidence of non-monotonic rate-level functions than those in the primary fields. Two further auditory fields, which were not tonotopically organized, were found on the anterior ectosylvian gyrus. Neurons in the more dorsal anterior area gave short-latency, transient responses to tones and were generally broadly tuned with a preference for high (>8 kHz) frequencies. Neurons in the other anterior area were frequently unresponsive to tones, but often responded vigorously to broadband noise. The presence of both tonotopic and non-tonotopic auditory cortical fields indicates that the organization of ferret auditory cortex is comparable to that seen in other mammals.

Scientific discoveries often hinge on synthesizing decades of research, a task that potentially outstrips human information processing capacities. Large language models (LLMs) offer a solution. LLMs trained on the vast scientific literature could potentially integrate noisy yet interrelated findings to forecast novel results better than human experts. Here, to evaluate this possibility, we created BrainBench, a forward-looking benchmark for predicting neuroscience results. We find that LLMs surpass experts in predicting experimental outcomes. BrainGPT, an LLM we tuned on the neuroscience literature, performed better yet. Like human experts, when LLMs indicated high confidence in their predictions, their responses were more likely to be correct, which presages a future where LLMs assist humans in making discoveries. Our approach is not neuroscience specific and is transferable to other knowledge-intensive endeavours.

How and where in the brain audio-visual signals are bound to create multimodal objects remains unknown. One hypothesis is that temporal coherence between dynamic multisensory signals provides a mechanism for binding stimulus features across sensory modalities. Here, we report that when the luminance of a visual stimulus is temporally coherent with the amplitude fluctuations of one sound in a mixture, the representation of that sound is enhanced in auditory cortex. Critically, this enhancement extends to include both binding and non-binding features of the sound. We demonstrate that visual information conveyed from visual cortex via the phase of the local field potential is combined with auditory information within auditory cortex. These data provide evidence that early cross-sensory binding provides a bottom-up mechanism for the formation of cross-sensory objects and that one role for multisensory binding in auditory cortex is to support auditory scene analysis.

Because we can perceive the pitch, timbre, and spatial location of a sound source independently, it seems natural to suppose that cortical processing of sounds might separate out spatial from nonspatial attributes. Indeed, recent studies support the existence of anatomically segregated “what” and “where” cortical processing streams. However, few attempts have been made to measure the responses of individual neurons in different cortical fields to sounds that vary simultaneously across spatial and nonspatial dimensions. We recorded responses to artificial vowels presented in virtual acoustic space to investigate the representations of pitch, timbre, and sound source azimuth in both core and belt areas of ferret auditory cortex. A variance decomposition technique was used to quantify the way in which altering each parameter changed neural responses. Most units were sensitive to two or more of these stimulus attributes. Although indicating that neural encoding of pitch, location, and timbre cues is distributed across auditory cortex, significant differences in average neuronal sensitivity were observed across cortical areas and depths, which could form the basis for the segregation of spatial and nonspatial cues at higher cortical levels. Some units exhibited significant nonlinear interactions between particular combinations of pitch, timbre, and azimuth. These interactions were most pronounced for pitch and timbre and were less commonly observed between spatial and nonspatial attributes. Such nonlinearities were most prevalent in primary auditory cortex, although they tended to be small compared with stimulus main effects.