Artificial intelligence (AI) is rapidly becoming one of the key technologies of this century. The majority of results in AI thus far have been achieved using deep neural networks trained with a learning algorithm called error backpropagation, always considered biologically implausible. To this end, recent works have studied learning algorithms for deep neural networks inspired by the neurosciences. One such theory, called predictive coding (PC), has shown promising properties that make it potentially valuable for the machine learning community: it can model information processing in different areas of the brain, can be used in control and robotics, has a solid mathematical foundation in variational inference, and performs its computations asynchronously. Inspired by such properties, works that propose novel PC-like algorithms are starting to be present in multiple sub-fields of machine learning and AI at large. Here, we survey such efforts by first providing a broad overview of the history of PC to provide common ground for the understanding of the recent developments, then by describing current efforts and results, and concluding with a large discussion of possible implications and ways forward.

Effective teaching necessitates adapting pedagogical strategies to the inherent diversity of students, encompassing variations in aptitude, learning styles, and personality, a critical challenge in education and teacher training. Large Language Models (LLMs) offer a powerful tool to simulate complex classroom dynamics, providing a controlled environment for exploring optimal teaching patterns. However, existing simulation frameworks often fall short by neglecting comprehensive student modeling beyond basic knowledge states and, more importantly, by lacking mechanisms for teachers to dynamically adapt their approach based on student feedback and collective performance. Addressing these limitations, we propose a simulation framework that integrates LLM-based diverse student agents with a self-evolving teacher agent. We use genetic algorithms to automatically tune and optimize the teacher’s pedagogical parameters based on simulated student performance, enabling the teacher agent to discover and refine teaching patterns tailored to specific class characteristics. Complementing this, we introduce Persona-RAG, a novel Retrieval-Augmented Generation method specifically designed for personalized knowledge retrieval in pedagogical contexts, allowing students to retrieve information as per their learning styles. We show how Persona-RAG remains competitive with standard RAG baselines in accurately retrieving relevant information while adding a touch of personalization for students. Crucially, we perform extensive experiments and highlight the different patterns learnt by the teacher agent while optimizing over classes with students of various learning styles. Our work presents a significant step towards creating adaptive educational technologies and improving teacher training through realistic, data-driven simulation.

This work presents a neuro-symbolic analysis of the learnability of Recurrent Neural Networks (RNNs) in classifying structured formal languages—specifically, **counter languages** and **Dyck languages**, which serve as canonical examples of context-free and mildly context-sensitive grammars. While prior studies have highlighted the expressive power of first-order (LSTM) and second-order (O2RNN) architectures within the Chomsky hierarchy, we challenge this perspective by shifting the focus from theoretical expressivity to *practical learnability under finite precision constraints*. Our results suggest that RNNs function more as finite-state machines than stack-based automata when implemented with realistic training regimes and embedding representations. We show that classification performance degrades sharply as structural similarities between positive and negative sequences increase—highlighting a core limitation in the RNN’s ability to internalize hierarchical structure without symbolic scaffolding. Interestingly, even simple linear classifiers built on top of RNN-derived embeddings outperform chance, underscoring the hidden representational capacity within learned states. To probe generalization, we train models on input lengths up to 40 and evaluate on lengths extending to 500, using 10 distinct seeds to measure statistical robustness. O2RNNs consistently demonstrate greater stability and generalization compared to LSTMs, particularly under varied initialization strategies. These findings expose the fragility of learned language representations and emphasize the role of architectural bias, initialization, and data sampling in determining what is truly learnable. Ultimately, our study reframes RNN learnability through the lens of *symbolic structure and computational constraints*, advocating for stronger formal criteria when assessing neural models’ capacity to reason over structured sequences. We argue that expressivity alone is insufficient—**stability, precision, and symbolic alignment** are essential for true neuro-symbolic generalization.

Prompt-based reasoning strategies such as Chain-of-Thought (CoT) and In-Context Learning (ICL) have become widely used for eliciting reasoning capabilities in Large Language Models (LLMs). However, these methods rely on fragile, implicit mechanisms often yielding inconsistent outputs across seeds, formats, or minor prompt variations making them fundamentally unreliable for tasks requiring stable, interpretable reasoning. In contrast, automata-based neuro-symbolic frameworks like RetoMaton offer a more structured and trustworthy alternative by grounding retrieval in symbolic memory with deterministic transitions. In this work, we extend RetoMaton by replacing its global datastore with a local, task-adaptive Weighted Finite Automaton (WFA), constructed directly from external domain corpora. This local automaton structure promotes robust, context-aware retrieval while preserving symbolic traceability and low inference overhead. Unlike prompting, which entangles context and memory in opaque ways, our approach leverages the explicit structure of WFAs to provide verifiable and modular retrieval behavior, making it better suited for domain transfer and interoperability. We evaluate this local RetoMaton variant on two pretrained LLMs LLaMA-3.2-1B and Gemma-3-1B-PT across three reasoning tasks: TriviaQA (reading comprehension), GSM8K (multi-step math), and MMLU (domain knowledge). Compared to the base model and prompting-based methods, augmenting these setups with local RetoMaton consistently improves performance while enabling transparent and reproducible retrieval dynamics. Our results highlight a promising shift toward trustworthy, symbolic reasoning in modern LLMs via lightweight, automaton-guided memory.

Surgical pathology reports are foundational to cancer diagnosis, staging, and treatment planning, yet they remain among the most complex and opaque documents in clinical practice. Their dense language, inconsistent structure, and use of domain-specific terminology pose significant barriers not only to patients but also to clinical teams seeking standardized, guideline-concordant care. In this study, we investigate the use of chain-of-thought (CoT) prompting with large language models (LLMs) to improve the interpretability and utility of pathology reports in endometrial cancer―the most common gynecologic malignancy in developed countries. Our approach targets three key objectives: (1) evaluating whether CoT-guided LLMs can extract and infer clinically relevant features from unstructured pathology narratives with high accuracy; (2) generating dual-layered summaries tailored for both oncology teams and patients, preserving diagnostic rigor while enhancing accessibility; and (3) assessing the impact of expert-in-the-loop feedback in refining reasoning quality and staging alignment with AJCC and FIGO guidelines. To benchmark model performance and evaluate clinical relevance, we utilized annotated cases from The Cancer Genome Atlas (TCGA) Endometrial Carcinoma Study, leveraging its curated pathology and genomic data to assess extraction fidelity and staging concordance. We demonstrate that GPT-4o, when augmented with CoT prompts, reduces optical character recognition (OCR) word error rate from 11.2% to 6.7% and improves character accuracy from 94.1% to 96.5%. Information extraction fidelity improves substantially, with histologic type classification rising from 82.0% to 94.1% and staging accuracy from 63.1% to 85.6%. These findings suggest that CoT-guided LLMs offer a scalable framework for enhancing diagnostic communication, improving clinical workflow efficiency, and supporting precision oncology through consistent, interpretable, and patient-centered information delivery.

Training deep neural networks on large-scale datasets requires significant hardware resources whose costs (even on cloud platforms) put them out of reach of smaller organizations, groups, and individuals. Backpropagation (backprop), the workhorse for training these networks, is an inherently sequential process that is difficult to parallelize. Furthermore, researchers must continually develop various specialized techniques, such as particular weight initializations and enhanced activation functions, to ensure stable parameter optimization. Our goal is to seek an effective, neuro-biologically plausible alternative to backprop that can be used to train deep networks. In this paper, we propose a backprop-free procedure, recursive local representation alignment, for training large-scale architectures. Experiments with residual networks on CIFAR-10 and the large benchmark, ImageNet, show that our algorithm generalizes as well as backprop while converging sooner due to weight updates that are parallelizable and computationally less demanding. This is empirical evidence that a backprop-free algorithm can scale up to larger datasets.

In this article, we propose a backpropagation-free approach to robotic control through the neuro-cognitive computational framework of neural generative coding (NGC), designing an agent completely built from predictive processing circuits that facilitate dynamic, online learning from sparse rewards, embodying the principles of planning-as-inference. Concretely, we craft an adaptive agent system, which we call active predictive coding (ActPC), that balances an internally-generated epistemic signal (meant to encourage intelligent exploration) with an internally-generated instrumental signal (meant to encourage goal-seeking behavior) to learn how to control various simulated robotic systems as well as a complex robotic arm using a realistic simulator, i.e., the Surreal Robotics Suite, for the block lifting task and the can pick-and-place problem. Notably, our results demonstrate that the proposed ActPC agent performs well in the face of sparse (extrinsic) reward signals and is competitive with or outperforms several powerful backpropagation-based reinforcement learning approaches.

In humans, perceptual awareness facilitates the fast recognition and extraction of information from sensory input. This awareness largely depends on how the human agent interacts with the environment. In this work, we propose active neural generative coding, a computational framework for learning action-driven generative models without backpropagation of errors (backprop) in dynamic environments. Specifically, we develop an intelligent agent that operates even with sparse rewards, drawing inspiration from the cognitive theory of planning as inference. We demonstrate on several simple control problems that our framework performs competitively with deep Q-learning. The robust performance of our agent offers promising evidence that a backprop-free approach for neural inference and learning can drive goal-directed behavior.

Recent advances in deep learning have led to superhuman performance across a variety of applications. Recently, these methods have been successfully employed to improve the rate-distortion performance in the task of image compression. However, current methods either use additional post-processing blocks on the decoder end to improve compression or propose an end-to-end compression scheme based on heuris-tics. For the majority of these, the trained deep neural networks (DNNs) are not compatible with standard encoders and would be difficult to deploy on personal com-puters and cellphones. In light of this, we propose a system that learns to improve the encoding performance by enhancing its internal neural representations on both the encoder and decoder ends, an approach we call Neural JPEG. We propose frequency domain pre-editing and post-editing methods to optimize the distribution of the DCT coefficients at both encoder and decoder ends in order to improve the stan-dard compression (JPEG) method. Moreover, we design and integrate a scheme for jointly learning quantization tables within this hybrid neural compression framework. In summary, our contributions are as follows:

In lifelong learning systems based on artificial neural networks, one of the biggest obstacles is the inability to retain old knowledge as new information is encountered. This phenomenon is known as catastrophic forgetting. In this paper, we propose a new kind of connectionist architecture, the Sequential Neural Coding Network, that is robust to forgetting when learning from streams of data points and, unlike networks of today, does not learn via the popular back-propagation of errors. Grounded in the neurocognitive theory of predictive coding, our model adapts its synapses in a biologically-plausible fashion while another neural system learns to direct and control this cortex-like structure, mimicking some of the task-executive control functionality of the basal ganglia. In our experiments, we demonstrate that our self-organizing system experiences significantly less forgetting compared to standard neural models, outperforming a swath of previously proposed methods, including rehearsal/data buffer-based methods, on both standard (SplitMNIST, Split Fashion MNIST, etc.) and custom benchmarks even though it is trained in a stream-like fashion. Our work offers evidence that emulating mechanisms in real neuronal systems, e.g., local learning, lateral competition, can yield new directions and possibilities for tackling the grand challenge of lifelong machine learning.

Automated mathematical reasoning is a challenging problem that requires an agent to learn algebraic patterns that contain long-range dependencies. Two particular tasks that test this type of reasoning are (1)mathematical equation verification,which requires determining whether trigonometric and linear algebraic statements are valid identities or not, and (2)equation completion, which entails filling in a blank within an expression to make it true. Solving these tasks with deep learning requires that the neural model learn how to manipulate and compose various algebraic symbols, carrying this ability over to previously unseen expressions. Artificial neural net-works, including recurrent networks and transformers, struggle to generalize on these kinds of difficult compositional problems, often exhibiting poor extrapolation performance.In contrast, recursive neural networks (recursive-NNs) are,theoretically, capable of achieving better extrapolation due to their tree-like design but are very difficult to optimize as the depth of their underlying tree structure increases. To over-come this, we extend recursive-NNs to utilize multiplicative,higher-order synaptic connections and, furthermore, to learn to dynamically control and manipulate an external memory.We argue that this key modification gives the neural system the ability to capture powerful transition functions for each possible input. We demonstrate the effectiveness of our pro-posed higher-order, memory-augmented recursive-NN models on two challenging mathematical equation tasks, showing improved extrapolation, stable performance, and faster convergence. We show that our models achieve 1.53% average improvement over current state-of-the-art methods in equation verification and achieve 2.22% top-1 average accuracy and 2.96% top-5 average accuracy for equation completion.

Recent advances in deep learning have resulted in image compression algorithms that outperform JPEG and JPEG 2000 on the standard Kodak benchmark. However, they are slow to train (due to backprop-through-time) and, to the best of our knowledge, have not been systematically evaluated on a large variety of datasets. In this paper, we perform the first large scale comparison of recent state-of-the-art hybrid neural compression algorithms, while exploring the effects alternative training strategies (when applicable). The hybrid recurrent neural decoder is a former state-of-the-art model (recently overtaken by a Google model) that can be trained using backprop-through-time (BPTT) or with alternative algorithms like sparse attentive backtracking (SAB), unbiased online recurrent optimization (UORO), and real time recurrent learning (RTRL). We compare these training alternatives along with the Google models (GOOG and E2E) on 6 benchmark datasets. Surprisingly, we found that the model trained with SAB performs the better (outperforming even BPTT), resulting in faster convergence and better peak signal-to-noise ratio.

For lossy image compression, we develop a neural-based system which learns a nonlinear estimator for decoding from quantized representations. The system links two recurrent networks that "help" each other reconstruct the same target image patches using complementary portions of the spatial context, communicating with each other via gradient signals. This dual agent system builds upon prior work that proposed an iterative refinement algorithm for recurrent neural network (RNN) based decoding. Our approach works with any neural or non-neural encoder. Our system progressively reduces image patch reconstruction error over a fixed number of steps. Experiments with variations of RNN memory cells show that our system consistently creates lower distortion images of higher perceptual quality compared to other approaches. Specifically, on the Kodak Lossless True Color Image Suite, we observe gains of 1:64 decibel (dB) over JPEG, a 1:46 dB over JPEG2000, a 1:34 dB over the GOOG neural baseline, 0:36 over E2E (a modern competitive neural compression model), and 0:37 over a single iterative neural decoder.

Temporal models based on recurrent neural networks have proven to be quite powerful in a wide variety of applications, including language modeling and speech processing. However, training these models often relies on backpropagation through time (BPTT), which entails unfolding the network over many time steps, making the process of conducting credit assignment considerably more challenging. Furthermore, the nature of backpropagation itself does not permit the use of nondifferentiable activation functions and is inherently sequential, making parallelization of the underlying training process difficult. Here, we propose the parallel temporal neural coding network (P-TNCN), a biologically inspired model trained by the learning algorithm we call local representation alignment. It aims to resolve the difficulties and problems that plague recurrent networks trained by BPTT. The architecture requires neither unrolling in time nor the derivatives of its internal activation functions. We compare our model and learning procedure with other BPTT alternatives (which also tend to be computationally expensive), including real-time recurrent learning, echo state networks, and unbiased online recurrent optimization. We show that it outperforms these on-sequence modeling benchmarks such as Bouncing MNIST, a new benchmark we denote as Bouncing NotMNIST, and Penn Treebank. Notably, our approach can, in some instances, outperform full BPTT as well as variants such as sparse attentive backtracking. Significantly, the hidden unit correction phase of P-TNCN allows it to adapt to new data sets even if its synaptic weights are held fixed (zero-shot adaptation) and facilitates retention of prior generative knowledge when faced with a task sequence. We present results that show the P-TNCN's ability to conduct zero-shot adaptation and online continual sequence modeling.

Finding biologically plausible alternatives to back-propagation of errors is a fundamentally important challenge in artificial neural network research. In this paper, we propose a learning algorithm called error-driven Local Representation Alignment (LRA-E), which has strong connections to predictive coding, a theory that offers a mechanistic way of describing neurocomputational machinery. In addition, we propose an improved variant of Difference Target Propagation, another procedure that comes from the same family of algorithms as LRA-E. We compare our procedures to several other biologicallymotivated algorithms, including two feedback alignment algorithms and Equilibrium Propagation. In two benchmarks, we find that both of our proposed algorithms yield stable performance and strong generalization compared to other competing back-propagation alternatives when training deeper, highly nonlinear networks, with LRA-E performing the best overall.

We examine the benefits of visual context in training neural language models to perform next-word prediction.A multi-modal neural architecture is introduced that outperform its equivalent trained on language alone with a 2% decrease in perplexity, even when no visual context is available at test.Fine-tuning the embeddings of a pre-trained state-of-theart bidirectional language model (BERT) in the language modeling framework yields a 3.5% improvement.The advantage for training with visual context when testing without is robust across different languages (English, German and Spanish) and different models (GRU, LSTM, ∆-RNN, as well as those that use BERT embeddings).Thus, language models perform better when they learn like a baby, i.e, in a multi-modal environment.This finding is compatible with the theory of situated cognition: language is inseparable from its physical context.

We propose novel neural temporal models for predicting and synthesizing human motion, achieving state-of-the-art in modeling long-term motion trajectories while being competitive with prior work in short-term prediction and requiring significantly less computation. Key aspects of our proposed system include: 1) a novel, two-level processing architecture that aids in generating planned trajectories, 2) a simple set of easily computable features that integrate derivative information, and 3) a novel multi-objective loss function that helps the model to slowly progress from simple next-step prediction to the harder task of multi-step, closed-loop prediction. Our results demonstrate that these innovations improve the modeling of long-term motion trajectories. Finally, we propose a novel metric, called Normalized Power Spectrum Similarity (NPSS), to evaluate the long-term predictive ability of motion synthesis models, complementing the popular mean-squared error (MSE) measure of Euler joint angles over time. We conduct a user study to determine if the proposed NPSS correlates with human evaluation of long-term motion more strongly than MSE and find that it indeed does. We release code and additional results (visualizations) for this paper at: https://github.com/cr7anand/neural_temporal_models.

For lossy image compression systems, we develop an algorithm, iterative refinement, to improve the decoder's reconstruction compared to standard decoding techniques. Specifically, we propose a recurrent neural network approach for nonlinear, iterative decoding. Our decoder, which works with any encoder, employs self-connected memory units that make use of causal and non-causal spatial context information to progressively reduce reconstruction error over a fixed number of steps. We experiment with variants of our estimator and find that iterative refinement consistently creates lower distortion images of higher perceptual quality compared to other approaches. Specifically, on the Kodak Lossless True Color Image Suite, we observe as much as a 0.871 decibel (dB) gain over JPEG, a 1.095 dB gain over JPEG 2000, and a 0.971 dB gain over a competitive neural model.