Narsis A. Kiani

Srijith Radhakrishnan, Chao-Han Huck Yang, Sumeer Ahmad Khan et al.

We introduce a new cross-modal fusion technique designed for generative error correction in automatic speech recognition (ASR). Our methodology leverages both acoustic information and external linguistic representations to generate accurate speech transcription contexts. This marks a step towards a fresh paradigm in generative error correction within the realm of n-best hypotheses. Unlike the existing ranking-based rescoring methods, our approach adeptly uses distinct initialization techniques and parameter-efficient algorithms to boost ASR performance derived from pre-trained speech and text models. Through evaluation across diverse ASR datasets, we evaluate the stability and reproducibility of our fusion technique, demonstrating its improved word error rate relative (WERR) performance in comparison to n-best hypotheses by relatively 37.66%. To encourage future research, we have made our code and pre-trained models open source at https://github.com/Srijith-rkr/Whispering-LLaMA.

Srijith Radhakrishnan, Chao-Han Huck Yang, Sumeer Ahmad Khan et al.

In this work, we explore Parameter-Efficient-Learning (PEL) techniques to repurpose a General-Purpose-Speech (GSM) model for Arabic dialect identification (ADI). Specifically, we investigate different setups to incorporate trainable features into a multi-layer encoder-decoder GSM formulation under frozen pre-trained settings. Our architecture includes residual adapter and model reprogramming (input-prompting). We design a token-level label mapping to condition the GSM for Arabic Dialect Identification (ADI). This is challenging due to the high variation in vocabulary and pronunciation among the numerous regional dialects. We achieve new state-of-the-art accuracy on the ADI-17 dataset by vanilla fine-tuning. We further reduce the training budgets with the PEL method, which performs within 1.86% accuracy to fine-tuning using only 2.5% of (extra) network trainable parameters. Our study demonstrates how to identify Arabic dialects using a small dataset and limited computation with open source code and pre-trained models.

Haoling Zhang, Chao-Han Huck Yang, Hector Zenil et al.

NeuroEvolution is one of the most competitive evolutionary learning frameworks for designing novel neural networks for use in specific tasks, such as logic circuit design and digital gaming. However, the application of benchmark methods such as the NeuroEvolution of Augmenting Topologies (NEAT) remains a challenge, in terms of their computational cost and search time inefficiency. This paper advances a method which incorporates a type of topological edge coding, named Reverse Encoding Tree (RET), for evolving scalable neural networks efficiently. Using RET, two types of approaches -- NEAT with Binary search encoding (Bi-NEAT) and NEAT with Golden-Section search encoding (GS-NEAT) -- have been designed to solve problems in benchmark continuous learning environments such as logic gates, Cartpole, and Lunar Lander, and tested against classical NEAT and FS-NEAT as baselines. Additionally, we conduct a robustness test to evaluate the resilience of the proposed NEAT algorithms. The results show that the two proposed strategies deliver improved performance, characterized by (1) a higher accumulated reward within a finite number of time steps; (2) using fewer episodes to solve problems in targeted environments, and (3) maintaining adaptive robustness under noisy perturbations, which outperform the baselines in all tested cases. Our analysis also demonstrates that RET expends potential future research directions in dynamic environments. Code is available from https://github.com/HaolingZHANG/ReverseEncodingTree.

Abdulrahman Ibraheem, Narsis Kiani, Jesper Tegner

This paper studies the usefulness of incorporating path information in predicting chemical properties from molecular graphs, in the domain of QSAR (Quantitative Structure-Activity Relationship). Towards this, we developed a GNN-style model which can be toggled to operate in one of two modes: a non-degenerate mode which incorporates path information, and a degenerate mode which leaves out path information. Thus, by comparing the performance of the non-degenerate mode versus the degenerate mode on relevant QSAR datasets, we were able to directly assess the significance of path information on those datasets. Our results corroborate previous works, by suggesting that the usefulness of path information is datasetdependent. Unlike previous studies however, we took the very first steps towards building a model that could predict upfront whether or not path information would be useful for a given dataset at hand. Moreover, we also found that, albeit its simplicity, the degenerate mode of our model yielded rather surprising results, which outperformed more sophisticated SOTA models in certain cases.

Santiago Hernández-Orozco, Hector Zenil, Jürgen Riedel et al.

We show how complexity theory can be introduced in machine learning to help bring together apparently disparate areas of current research. We show that this new approach requires less training data and is more generalizable as it shows greater resilience to random attacks. We investigate the shape of the discrete algorithmic space when performing regression or classification using a loss function parametrized by algorithmic complexity, demonstrating that the property of differentiation is not necessary to achieve results similar to those obtained using differentiable programming approaches such as deep learning. In doing so we use examples which enable the two approaches to be compared (small, given the computational power required for estimations of algorithmic complexity). We find and report that (i) machine learning can successfully be performed on a non-smooth surface using algorithmic complexity; (ii) that parameter solutions can be found using an algorithmic-probability classifier, establishing a bridge between a fundamentally discrete theory of computability and a fundamentally continuous mathematical theory of optimization methods; (iii) a formulation of an algorithmically directed search technique in non-smooth manifolds can be defined and conducted; (iv) exploitation techniques and numerical methods for algorithmic search to navigate these discrete non-differentiable spaces can be performed; in application of the (a) identification of generative rules from data observations; (b) solutions to image classification problems more resilient against pixel attacks compared to neural networks; (c) identification of equation parameters from a small data-set in the presence of noise in continuous ODE system problem, (d) classification of Boolean NK networks by (1) network topology, (2) underlying Boolean function, and (3) number of incoming edges.

Rise Ooi, Chao-Han Huck Yang, Pin-Yu Chen et al.

Networks are fundamental building blocks for representing data, and computations. Remarkable progress in learning in structurally defined (shallow or deep) networks has recently been achieved. Here we introduce evolutionary exploratory search and learning method of topologically flexible networks under the constraint of producing elementary computational steady-state input-output operations. Our results include; (1) the identification of networks, over four orders of magnitude, implementing computation of steady-state input-output functions, such as a band-pass filter, a threshold function, and an inverse band-pass function. Next, (2) the learned networks are technically controllable as only a small number of driver nodes are required to move the system to a new state. Furthermore, we find that the fraction of required driver nodes is constant during evolutionary learning, suggesting a stable system design. (3), our framework allows multiplexing of different computations using the same network. For example, using a binary representation of the inputs, the network can readily compute three different input-output functions. Finally, (4) the proposed evolutionary learning demonstrates transfer learning. If the system learns one function A, then learning B requires on average less number of steps as compared to learning B from tabula rasa. We conclude that the constrained evolutionary learning produces large robust controllable circuits, capable of multiplexing and transfer learning. Our study suggests that network-based computations of steady-state functions, representing either cellular modules of cell-to-cell communication networks or internal molecular circuits communicating within a cell, could be a powerful model for biologically inspired computing. This complements conceptualizations such as attractor based models, or reservoir computing.

Hector Zenil, Narsis A. Kiani, Allan A. Zea et al.

Complex data usually results from the interaction of objects produced by different generating mechanisms. Here we introduce a universal, unsupervised and parameter-free model-oriented approach, based upon the seminal concept of algorithmic probability, that decomposes an observation into its most likely algorithmic generative sources. Our approach uses a causal calculus to infer model representations. We demonstrate its ability to deconvolve interacting mechanisms regardless of whether the resultant objects are strings, space-time evolution diagrams, images or networks. While this is mostly a conceptual contribution and a novel framework, we provide numerical evidence evaluating the ability of our methods to separate data from observations produced by discrete dynamical systems such as cellular automata and complex networks. We think that these separating techniques can contribute to tackling the challenge of causation, thus complementing other statistically oriented approaches.

Santiago Hernández-Orozco, Narsis A. Kiani, Hector Zenil

Natural selection explains how life has evolved over millions of years from more primitive forms. The speed at which this happens, however, has sometimes defied formal explanations when based on random (uniformly distributed) mutations. Here we investigate the application of a simplicity bias based on a natural but algorithmic distribution of mutations (no recombination) in various examples, particularly binary matrices in order to compare evolutionary convergence rates. Results both on synthetic and on small biological examples indicate an accelerated rate when mutations are not statistical uniform but \textit{algorithmic uniform}. We show that algorithmic distributions can evolve modularity and genetic memory by preservation of structures when they first occur sometimes leading to an accelerated production of diversity but also population extinctions, possibly explaining naturally occurring phenomena such as diversity explosions (e.g. the Cambrian) and massive extinctions (e.g. the End Triassic) whose causes are currently a cause for debate. The natural approach introduced here appears to be a better approximation to biological evolution than models based exclusively upon random uniform mutations, and it also approaches a formal version of open-ended evolution based on previous formal results. These results validate some suggestions in the direction that computation may be an equally important driver of evolution. We also show that inducing the method on problems of optimization, such as genetic algorithms, has the potential to accelerate convergence of artificial evolutionary algorithms.