Position Paper: An Inner Interpretability Framework for AI Inspired by Lessons from Cognitive Neuroscience

Inner Interpretability is a promising emerging field tasked with uncovering the inner mechanisms of AI systems, though how to develop these mechanistic theories is still much debated. Moreover, recent critiques raise issues that question its usefulness to advance the broader goals of AI. However, it has been overlooked that these issues resemble those that have been grappled with in another field: Cognitive Neuroscience. Here we draw the relevant connections and highlight lessons that can be transferred productively between fields. Based on these, we propose a general conceptual framework and give concrete methodological strategies for building mechanistic explanations in AI inner interpretability research. With this conceptual framework, Inner Interpretability can fend off critiques and position itself on a productive path to explain AI systems.

Memory in humans and deep language models: Linking hypotheses for model augmentation

The computational complexity of the self-attention mechanism in Transformer models significantly limits their ability to generalize over long temporal durations. Memory-augmentation, or the explicit storing of past information in external memory for subsequent predictions, has become a constructive avenue for mitigating this limitation. We argue that memory-augmented Transformers can benefit substantially from considering insights from the memory literature in humans. We detail an approach to integrating evidence from the human memory system through the specification of cross-domain linking hypotheses. We then provide an empirical demonstration to evaluate the use of surprisal as a linking hypothesis, and further identify the limitations of this approach to inform future research.

Successes and critical failures of neural networks in capturing human-like speech recognition

Natural and artificial audition can in principle evolve different solutions to a given problem. The constraints of the task, however, can nudge the cognitive science and engineering of audition to qualitatively converge, suggesting that a closer mutual examination would improve artificial hearing systems and process models of the mind and brain. Speech recognition - an area ripe for such exploration - is inherently robust in humans to a number transformations at various spectrotemporal granularities. To what extent are these robustness profiles accounted for by high-performing neural network systems? We bring together experiments in speech recognition under a single synthesis framework to evaluate state-of-the-art neural networks as stimulus-computable, optimized observers. In a series of experiments, we (1) clarify how influential speech manipulations in the literature relate to each other and to natural speech, (2) show the granularities at which machines exhibit out-of-distribution robustness, reproducing classical perceptual phenomena in humans, (3) identify the specific conditions where model predictions of human performance differ, and (4) demonstrate a crucial failure of all artificial systems to perceptually recover where humans do, suggesting a key specification for theory and model building. These findings encourage a tighter synergy between the cognitive science and engineering of audition.

Phonological (un)certainty weights lexical activation

Spoken word recognition involves at least two basic computations. First is matching acoustic input to phonological categories (e.g. /b/, /p/, /d/). Second is activating words consistent with those phonological categories. Here we test the hypothesis that the listener's probability distribution over lexical items is weighted by the outcome of both computations: uncertainty about phonological discretisation and the frequency of the selected word(s). To test this, we record neural responses in auditory cortex using magnetoencephalography, and model this activity as a function of the size and relative activation of lexical candidates. Our findings indicate that towards the beginning of a word, the processing system indeed weights lexical candidates by both phonological certainty and lexical frequency; however, later into the word, activation is weighted by frequency alone.